Intelligent Automated In Vitro Fertilization and Intracellular Sperm Injection Platform

Abstract

A method for robotic pipetting based on artificial intelligence for cell preparation is disclosed. The method includes training an artificial intelligence / machine learning system (AI / ML system) to classify images of a biological specimen. The method includes storing the classified images. The method includes receiving image objects from an imaging system which includes a microscope system, a camera system configured to receive imaging from the microscope system, and an illumination system configured to illuminate the biological material. The method includes processing the image objects using the AI / ML system to determine the presence of recoverable target cells. The method includes positioning a robotic pipette in a target physical orientation relative to the recoverable target cells. The method includes confirming the location of the robotic pipette in a target physical orientation using the AI / ML system. The method includes instructing the robotic pipette to retrieve the recoverable target cells from a target physical orientation.

Description

[Technical Field]

[0001] <Cross-reference of related applications> This application is a PCT international application, U.S. Provisional Patent Application No. 63 / 523,258, filed on 26 June 2023, the entire disclosure thereof is incorporated herein by reference. [Background technology]

[0002] Traditional in vitro fertilization (IVF) techniques have relied heavily on the assistance of human clinical embryologists and / or male pathologists to perform, evaluate, and / or respond to the requirements of the IVF process. This has resulted in expensive IVF interventions that limit access to IVF due to economic constraints, geographical and other limitations, and this is subject to fluctuations in inconsistencies across individual capabilities as well as healthcare settings, equipment, and expertise. [Overview of the Initiative] [Means for solving the problem]

[0003] This disclosure includes methods and systems for providing an intelligent automated system of interconnected robotic IVF modules, including intraplasmic sperm injection (ICSI) technology, which is supported by imaging and artificial intelligence / machine learning (AI / ML) processes for acquiring, storing, analyzing, performing, and reporting multiple materials, data, processes, actions, and results related to IVF / ICSI.

[0004] In some embodiments, the method may include an AI-based robotic pipetting operation for cell preparation, which includes training an artificial intelligence / machine learning system (AI / ML system) to classify images of a biological specimen, the classification being partially based on cellular material in the image that conforms to a target cell criterion, and storing the classified images. Image objects may be received from an imaging system which includes a microscope system, a camera system configured to receive imaging from the microscope system, and an illumination system configured to illuminate the biological material, the microscope system, camera system and illumination system being operably connected, and the image objects meet a predetermined threshold based partially on the classified images. Image objects may be processed to determine the presence of recoverable target cells using the AI / ML system, and the robotic pipette may be positioned in a target physical orientation relative to the recoverable target cells. The location of the robotic pipette is in a target physical orientation, which can be verified using the AI / ML system, and the robotic pipette can be instructed to retrieve recoverable target cells from the target physical orientation.

[0005] In some embodiments, the image object may be a still image, a time-lapse image, a video, or some other image type.

[0006] In some embodiments, a predetermined threshold can be the probability that at least a portion of the image objects belong to the classified image type in question.

[0007] In some embodiments, the classified image type in question may be an image of a classified cell type.

[0008] In some embodiments, the AI / ML system can be used to determine the presence of retrievable target cells, and includes repeatedly acquiring multiple image objects until at least one of the multiple image objects meets a predetermined threshold. The AI / ML system can repeatedly acquire multiple image objects, including changing the physical arrangement of the imaging system to acquire multiple image orientations from among the multiple image objects. The AI / ML system can repeatedly acquire multiple image objects, including changing the type of illumination used by the imaging system to acquire multiple image orientations within the multiple image objects.

[0009] In some embodiments, the illumination type can be polarized.

[0010] In some embodiments, the target cells that can be recovered can be single cells.

[0011] In some embodiments, the AI / ML system can use at least one neural network.

[0012] In some embodiments, each component of the imaging system may include at least one motor for positioning.

[0013] In some embodiments, the physical orientation of the target can be obtained in part based on rotating the stage in which the harvestable target cells are located.

[0014] In some embodiments, the method may include an AI-based robotic pipetting operation for sperm preparation, including arranging a container containing a semen sample and receiving an image object from an imaging system including a microscope system, a camera system and an illumination system. The liquefaction state of the semen sample can be determined at least partially based on the image object by using an artificial intelligence / machine learning system (AI / ML system) to verify the validity of the determined liquefaction state for sperm collection by comparing the determined semen liquefaction state with a stored liquefaction state standard. The sperm characteristics within the semen sample can be determined at least partially based on the image object, and the determined sperm characteristics and motility can be compared with a stored sperm standard using the AI / ML system to verify the validity of the determined sperm characteristics and motility for sperm collection. The sperm motility velocity can be determined at least partially based on the image object, and the determined sperm motility velocity can be compared with a stored motility standard using the AI / ML system to verify the validity of the determined sperm motility velocity for sperm separation. Once the determined liquefaction state, determined sperm characteristics and motility, and the validity of the determined sperm motility are confirmed, the robotic pipette can be instructed to collect sperm from the semen sample, and at least one sperm can be collected in the robotic pipette.

[0015] In some embodiments, the placement of the vessels can be performed at least partially by rotating the vessels using a staging system controlled by an AI / ML system.

[0016] In some embodiments, the velocity of sperm motility can be at least one of vertical motion, horizontal motion, or angular motion measurements.

[0017] In some embodiments, sperm characteristics can be at least one of sperm count, motility parameters, or morphological measurements.

[0018] In some embodiments, the AI / ML system can use at least one neural network.

[0019] In some embodiments, the sperm collected by the robotic pipette can be multiple sperm.

[0020] In some embodiments, a method of robotic pipetting based on artificial intelligence for sperm preparation, comprising using an artificial intelligence / machine learning (AI / ML system) system to place a vessel containing a semen sample on a staging mechanism, the staging mechanism including at least one motor for placing the vessel in a specified direction. The AI / ML system can be used to confirm the position of the vessel and can instruct the robotic pipette to place the pipette tip within the semen sample at a position specified by the AI / ML system. The robotic pipette can be used to aspirate and mix the semen sample using the robotic pipette. An image object can be received from an imaging system including a microscope system, a camera system, and a lighting system. The AI / ML system can be used to determine the semen liquefaction state of the semen sample based at least in part on the image object. The determined semen liquefaction state can be compared with a predetermined liquefaction threshold using the AI / ML system to confirm the validity of the determined semen liquefaction state for sperm collection. The robotic pipette can be used to mix the semen sample according to the liquefaction state evaluation and can be used to obtain at least one sperm from the semen sample at a position specified by the AI / ML system.

[0021] In some embodiments, the specified direction can be the degree of inclination, the degree of rotation, the speed of movement, and can be determined algorithmically by the AI / ML system.

[0022] In some embodiments, the image object can be a still image, a time-lapse image, a video, or some other image type.

[0023] In some embodiments, the aspiration and mixing of semen is repeatedly performed using an AI / ML system and an imaging system until a predetermined liquefaction threshold is met by at least one of the plurality of image objects, generating a plurality of image objects.

[0024] In some embodiments, the mixing can include the addition of a culture medium containing a liquefaction reducing agent.

[0025] In some embodiments, the mixing may not include the addition of a culture medium containing a liquefaction reducing agent.

[0026] In some embodiments, the semen sample can be placed within a layer of culture medium within the vessel.

[0027] [[ID=第十七条]] In some embodiments, the method may include an AI-based robotic pipetting operation for sperm preparation, which includes placing a first container containing a semen sample within a staging mechanism, and receiving an image object showing the sperm in the semen sample from an imaging system including a microscope system, a camera system, and an illumination system. Multiple sperm within the image object can be graded using an artificial intelligence system / machine learning (AI / ML system) according to the degree to which each of the multiple sperm corresponds to a sperm quality standard. The AI / ML system can select high-grade sperm from the graded multiple sperm. The robotic pipette can collect and isolate the high-grade sperm into a second container, and the AI / ML system can be used to move the staging mechanism to position the second container in a designated location for inspection by the imaging system. Using the AI / ML system, an immobilization device can be instructed to block the tail function of high-grade sperm, which reduces the mobility of high-grade sperm. The robotic pipette can then be guided to position the pipette tip in a second container at a location specified by the AI / ML system to collect the high-grade sperm, and the high-grade sperm can be collected from the second container at the location specified by the AI / ML system using the robotic pipette.

[0028] In some embodiments, the first image object may be at least one of a single image, a time-series image, or a video.

[0029] In some embodiments, a decrease in the motility of high-grade sperm can be automatically measured by the AI / ML system.

[0030] In some embodiments, an AI / ML system can be used to determine sufficient immobilization by comparing an automated measurement of the decrease in mobility with a predetermined threshold.

[0031] In some embodiments, high-quality sperm can consist of multiple sperm.

[0032] In some embodiments, the immobilization device can be a laser, a pipette, or a piezoelectric actuator.

[0033] In some embodiments, a method of robotic pipetting operation based on artificial intelligence for sperm preparation, comprising placing a container containing a semen sample within a staging mechanism. The robotic pipette is capable of producing at least two droplets in a dish on the staging mechanism, and the robotic pipette can be used to connect at least two droplets to a culture medium pathway according to a programmable design commanded by an artificial intelligence / machine learning system (AI / ML system). The robotic pipette can deposit a certain amount of sperm from the semen sample into one of at least two droplets, and the AI / ML system can be used to optically scan a certain amount of sperm to generate a first image object and a second image object using an imaging system including a microscope system, a camera system and an illumination system, the creation of the first and second image objects being separated by a specified period. The AI / ML system can compare at least one sperm component of a certain amount of sperm in a first image object with at least one sperm component of a certain amount of sperm in a second image object to calculate the real-time motility index of multiple sperm from a certain amount of sperm, at least partially based on the change in sperm position between the first and second image objects, and can rank the multiple sperm at least partially based on the real-time motility index.

[0034] In some embodiments, the first and second image objects may be still images, time-lapse images, videos, or any other image type.

[0035] In some embodiments, a certain amount of sperm can be obtained without seminal plasma components.

[0036] In some embodiments, at least one of the two droplets may consist of untreated seminal plasma that does not contain culture medium.

[0037] In some embodiments, the semen sample can be diluted by mixing the semen sample with a culture medium.

[0038] In some embodiments, the optimal sperm preparation concentration can be maintained in at least two droplets by blocking the culture medium pathway, at least partially based on automatically blocking the culture medium pathway using the pipette tip.

[0039] In some embodiments, the imaging system can use autofocus.

[0040] In some embodiments, multiple robotic pipettes can be positioned at multiple angles relative to the container position specified by the AI / ML system.

[0041] In some embodiments, at least one sperm component may be a tail, acrosome, head, or midpiece.

[0042] In some embodiments, the comparison can be performed at least in part on an algorithmic comparison of at least one pixel change in a second image object relative to a first image object.

[0043] In some embodiments, the comparison can be performed sequentially.

[0044] In some embodiments, an artificial intelligence-based robotic pipetting method for sperm preparation comprises using an artificial intelligence / machine learning system (AI / ML system) to optically scan a semen sample and generate a first image object and a second image object using an imaging system including a microscope system, a camera system and an illumination system, wherein the creation of the first and second image objects exhibits the same focal space and their creations are separated by a specified period. The first and second image objects can be compared using the AI / ML system, which at least partially algorithmically measures at least one pixel change between the first and second image objects by estimating multiple real-time sperm motility indices for at least one subset of sperm present in the first and second image objects. The AI / ML system can identify at least one sperm having real-time sperm motility indicators that meet a predetermined threshold, control the imaging system to automatically focus on at least one sperm, position a robotic pipette in the target physical orientation relative to at least one sperm, and use the robotic pipette to obtain at least one sperm from the semen sample at the location specified by the AI / ML system.

[0045] In some embodiments, the semen sample may be untreated seminal plasma without culture medium.

[0046] In some embodiments, the first image object and the second image object may be a time-series image or video.

[0047] In some embodiments, an artificial intelligence-based robotic vitrification method comprising: placing at least one oocyte in a buffer or CPA solution; using an AI / ML system to process image objects generated by an imaging system to determine the location of a retrievable target oocyte; positioning a robotic pipette in a target physical orientation relative to the retrievable target oocyte; and using an AI / ML system to confirm the position of the robotic pipette in a target physical orientation. The AI / ML system can initiate contact with the retrievable target oocyte and instruct the robotic pipette to apply negative pressure to fix the oocyte in the robotic pipette. A robotic microtool holder can lower the vitrification assembly or any other type of freezing device to a position close to the target physical orientation in the buffer or CPA solution dish. The robotic pipette can position the oocyte on the vitrification assembly or other freezing device in a dish containing droplets of buffer or CPA solution. The AI / ML system maintains the same relative physical arrangement of oocytes and vitrification assembly / freezing apparatus, and can hold the robotic pipette and robotic microtool holder stationary relative to each other to form the vitrification assembly. The system can robotically move the vitrification assembly through different solutions for a predetermined time in a predetermined order commanded by the AI / ML system, and can generate fluid crosslinks between two or more bound droplets containing a buffer or CPA until the vitrification assembly reaches a predetermined threshold concentration of the cryoprotectant.

[0048] In some embodiments, at least one oocyte can develop into an embryo.

[0049] In some embodiments, the oocyte can develop into multiple embryos.

[0050] In some embodiments, the image object may be a still image, a time-lapse image, a video image, or some other image type.

[0051] In some embodiments, negative pressure can be applied via an electric microinjector that is at least partially controlled by the AI / ML system.

[0052] In some embodiments, the robotic pipette may consist of multiple robotic pipettes.

[0053] In some embodiments, the physical orientation of the target can be obtained in part based on rotating the stage in which the retrievable target oocytes are located.

[0054] In some embodiments, the vitrified assembly can be a mitogen racket.

[0055] In some embodiments, the AI / ML system can use at least one neural network.

[0056] In some embodiments, an artificial intelligence-based robotic vitrification method comprising using an artificial intelligence system / machine learning system (AI / ML system) to issue automated commands to a robotic pipette to obtain oocytes from a container. The AI / ML system can guide the robotic pipette containing the oocytes into a droplet containing a freezing device or vitrification assembly held by a robotic microtool at least partially controlled by the AI / ML system, and can position the robotic pipette in close proximity to the vitrification assembly to create a vitrification assembly-pipette combination, the positioning being guided at least partially based on the analysis of at least one image object created by an imaging system using the AI / ML system, the imaging system comprising a microscope system, a camera system and an illumination system, and can robotically vary the concentration of cryoprotectant in the solution in which the vitrification assembly-pipette combination holds the oocytes, exposing the oocytes held by the vitrification assembly-pipette combination to gradually increasing concentrations of cryoprotectant until the oocytes reach a predetermined threshold concentration of the cryoprotectant.

[0057] In some embodiments, the oocyte may be an embryo or a group of embryos.

[0058] In some embodiments, the concentration of the cryoprotectant can be automatically evaluated using an AI / ML system and an imaging system.

[0059] In some embodiments, the vitrified assembly can be a mitogen racket.

[0060] In some embodiments, the exposure of oocytes to gradually increasing concentrations of cryoprotective agents can be automatically performed according to a predetermined exposure rate by robotic control of an AI / ML system and a vitrification assembly-pipette combination.

[0061] In some embodiments, the exposure rate of oocytes to gradually increasing concentrations of cryoprotective agent can be automatically determined in real time by an AI / ML system, at least partially based on the output of an imaging system.

[0062] In some embodiments, the image object can be a still image, a video, or some other image type.

[0063] In some embodiments, a method for COC recovery based on automated artificial intelligence, comprising using an artificial intelligence / machine learning system (AI / ML system) to optically scan a follicular fluid sample and generate image objects using an imaging system including a microscope system, a camera system and an illumination system. The image objects can be compared using the AI / ML system to a predetermined threshold, the predetermined threshold being an optical pattern having a probability of corresponding at least in part to a cumulus oocyte complex (COC). COCs in a follicular fluid sample can be identified at least in part based on image objects that satisfy the predetermined threshold. The location of COCs in a follicular fluid sample can be determined at least in part based on identifying a region of an image object corresponding to an optical pattern in an image object that satisfies the predetermined threshold. A robotic pipette can be instructed to collect the COCs at their location and can be collected within the robotic pipette.

[0064] In some embodiments, optical scanning of follicular fluid samples can be performed automatically, at least partially, using an AI / ML system and an imaging system.

[0065] In some embodiments, optical scanning of the follicular fluid sample can be performed automatically according to a predetermined scanning pattern at least partially specified by the AI / ML system.

[0066] In some embodiments, the illumination system can image the follicular fluid sample using multiple wavelengths of light.

[0067] In some embodiments, the image object can be a still image, a time-series image, a video, or some other image type.

[0068] In some embodiments, the region of the image object corresponding to the optical pattern can consist of multiple regions.

[0069] In some embodiments, the robotic pipette may consist of multiple robotic pipettes.

[0070] In some embodiments, the COC can consist of multiple COCs.

[0071] In some embodiments, a method for automated COC recovery based on artificial intelligence, comprising placing a container containing a follicular fluid sample on an electric stage of a microscope. The follicular fluid sample can be scanned robotically using an imaging system and an artificial intelligence / machine learning system (AI / ML system) to create an image object, the imaging system may include a microscope system, a camera system and an illumination system. The cumulus oocyte complex (COC) can be identified within the image object and its location, at least in part, based on comparing the image object with a predetermined threshold using the AI / ML system, the predetermined threshold being an optical pattern having a probability corresponding to a COC mass. The robotic pipette can be instructed to acquire the COC at that location. The robotic pipette can initiate aspiration by applying negative pressure to the COC and adjacent body fluids, at least partially based on the application of negative pressure to the COC and adjacent body fluids, the amount of body fluid aspirated being determined at least partially by the physical limits outside the COC detected within the image object, the robotic pipette can be removed from the container, the motorized stage can be moved to present a washing dish containing the processing medium in close proximity to the robotic pipette, the robotic pipette can be lowered into the washing area of ​​the washing dish, positive pressure can be applied to the robotic pipette to discharge the COC and adjacent body fluids into the medium, the volume discharged is controlled at least partially using an AI / ML system, and as a result, the positive pressure stops when the COC mass exits the robotic pipette.

[0072] In some embodiments, the container and washing dish can be a single common container.

[0073] In some embodiments, the microscope may be an inverted microscope or a movable microscope.

[0074] In some embodiments, the amount of body fluid aspirated can be determined at least partially by the AI / ML system.

[0075] In some embodiments, the identified COC can be multiple COCs.

[0076] In some embodiments, the robotic pipette may consist of multiple robotic pipettes.

[0077] In some embodiments, scanning of follicular fluid samples can be performed automatically, at least partially, using an AI / ML system and an imaging system.

[0078] In some embodiments, suction can be continued until the AI / ML system and imaging system confirm completion.

[0079] In some embodiments, confirmation of completion is performed using an AI / ML system and an imaging system to confirm, at least partially, that the COC mass has been sufficiently separated from other materials, based on a second image taken during or after aspiration.

[0080] In some embodiments, the method for automated artificial intelligence-based denudation comprises scanning a dish containing a cumulus-oocyte complex (COC) with a robot using an imaging system and an artificial intelligence / machine learning system (AI / ML system) to create an image object, wherein the imaging system comprises a microscope system, a camera system and an illumination system, and the dish contains an enzyme that removes cumulus cells from oocytes in the COC. The imaging system and the AI / ML system can identify the COC within the location of the COC, at least in part on comparing the image object with a predetermined threshold using the AI / ML system, where the predetermined threshold is an optical pattern having a probability corresponding to the COC mass. The robotic pipette can be instructed to collect the COC at the location of the COC, and the AI / ML system collects the COC into the robotic pipette, discharges the COC from the pipette and returns it to the dish, and repeats this until the AI / ML system confirms sufficient removal of cumulus and corona radiata cells from the COC using a second image object generated by the imaging system.

[0081] In some embodiments, the dish may include a plurality of indentations.

[0082] In some embodiments, the dish can be a tube.

[0083] In some embodiments, the microscope system may include a stereomicroscope, an inverted microscope, or a mobile microscope.

[0084] In some embodiments, the enzyme can be a plurality of enzymes.

[0085] In some embodiments, the enzyme can be a plurality of compounds.

[0086] In some embodiments, scanning can be performed automatically based on a predetermined pattern, at least partially using an AI / ML system and an imaging system.

[0087] In some embodiments, the scan can be automatically prepared to use a scan pattern, at least partially based on the AI / ML system and the imaging system, as well as the AI / ML system evaluation of at least one imaging system output.

[0088] In some embodiments, the robotic pipette may consist of multiple robotic pipettes.

[0089] In some embodiments, repeatedly collecting COC into a robotic pipette may include continuously reducing the inner diameter of the pipette used repeatedly by the robotic pipette.

[0090] In some embodiments, the COC mass can consist of multiple COC masses.

[0091] In some embodiments, a method for human oocyte preparation based on automated artificial intelligence comprises scanning a dish containing cumulus-oocyte complexes (COCs) with a robot using an imaging system and an artificial intelligence / machine learning system (AI / ML system) to create a first image object, wherein the imaging system comprises a microscope system, a camera system and an illumination system, and the dish contains an enzyme that removes cumulus cells from oocytes in the COC. The imaging system and the AI ​​system can identify the COCs and their locations within the image object, at least on the basis of comparing the first image object with a predetermined threshold using the AI / ML system, wherein the predetermined threshold is at least an optical pattern having a probability corresponding to a COC mass. A first robotic pipette can be instructed to collect the COCs at their locations. The AI / ML system can be used to determine a first aspirate volume of oocytes / body fluid into the first robotic pipette. The AI / ML system and the imaging system can be used to determine the presence of air bubbles in the first aspirate volume of oocytes / body fluid, and the first aspirate volume of oocytes / body fluid can be discharged into a receptor placed on a stage. The AI / ML system and imaging system can be used to guide the positioning of a second robotic pipette to aspirate a subset of the first aspirated volume of oocytes / body fluid, the second robotic pipette having a narrower inner diameter than the first robotic pipette, and multiple gradually narrowing robotic pipettes can be used to aspirate and discharge the volume of oocytes / body fluid until sufficient detachment is confirmed using a second image object by the AI / ML system and imaging system for at least one oocyte in the volume of oocytes / body fluid.

[0092] In some embodiments, the second robotic pipette may have the same inner diameter as the first robotic pipette.

[0093] In some embodiments, the stage can be an inverted microscope stage.

[0094] In some embodiments, scanning can be performed automatically according to a predetermined scanning pattern specified at least partially by the AI / ML system.

[0095] In some embodiments, the first and second image objects may be still images, time-series images, videos, or any other image type.

[0096] In some embodiments, an automated artificial intelligence-based method for ICSI is provided, comprising receiving at least one droplet containing an oocyte into a dish placed on a stage. The zona pellucida can be detected using an artificial intelligence / machine learning system (AI / ML system) and an imaging system, the imaging system including a microscope system, a camera system and an illumination system. The oocyte can be held using a robotic microtool. A robotic pipette can be lowered into the droplet. The AI / ML system and imaging system can be used to determine the region to hold the oocyte and position the robotic microtool in that region. The AI / ML system and imaging system can be used to instruct the robotic microtool to hold the oocyte in the robotic pipette by applying negative pressure. The AI / ML system and imaging system can be used to determine the target location where zona pellucida excision should be performed and move the oocyte to the target location. The AI / ML system and imaging system can be used to evaluate the thickness of the zona pellucida and determine the excision action, and a laser can be generated to excise a predetermined area and depth of the zona pellucida. The AI / ML system and imaging system can be used to define the injection path into the oocyte.

[0097] In some embodiments, the stage can be a microscope stage.

[0098] In some embodiments, the microscope stage can be placed in close proximity to an inverted microscope, stereomicroscope, mobile microscope, optical coherence tomography (OCT) system, optical coherence microscope system, or lensless microscope.

[0099] In some embodiments, the robotic microtool can be a robotic holding pipette.

[0100] In some embodiments, the robotic pipette can consist of multiple pipettes.

[0101] In some embodiments, the dish can be heated.

[0102] In some embodiments, the imaging system can generate at least one image having a mixed reality feel, which is a combination of a simulated image and a real image.

[0103] In some embodiments, the excision can be performed with the necessary strength and radius to excise an appropriate portion of the zona pellucida, thereby facilitating the entry of the ICSI needle through the zona pellucida without deforming the oocyte.

[0104] In some embodiments, the method for automated ICSI based on artificial intelligence includes receiving an oocyte in a dish placed on a microscope stage. An artificial intelligence / machine learning system (AI / ML system) and an imaging system can be used to apply a combination of positive and negative pressure inside a robotic ICSI needle to position sperm at a desired location. The robotic ICSI needle can be advanced into the oocyte at a predetermined controlled speed and stopped when it reaches the end of a path specified by the AI / ML system, and piezoelectric pulses can be used to break the egg membrane (egg cell membrane). The AI / ML system and imaging system can be used to determine whether piezoelectric pulses are necessary to break the membrane and how many pulses are needed. The AI / ML system and imaging system can be used to apply positive pressure to the robotic ICSI needle to deposit sperm into the oocyte. The AI / ML system and imaging system can be used to confirm that the sperm are outside the needle and to move the needle out of the oocyte, and can be used to apply positive pressure to a holding pipette until the oocyte is released into a droplet.

[0105] In some embodiments, the egg membrane can be destroyed mechanically.

[0106] In some embodiments, positive pressure within the ICSI needle can push sperm towards the tip of the ICSI needle in preparation for injection into the cytoplasm, reducing the amount of exogenous culture medium injected into the oocyte.

[0107] In some embodiments, the desired location can be determined in part based on the use of conventional microscopy, optical coherence tomography, optical coherence microscopy, or three-dimensional simulation of oocyte morphology.

[0108] In some embodiments, a method for oocyte identification based on automated artificial intelligence using optical coherence tomography (OCT) is provided, comprising positioning an OCT imaging system head in close proximity to a biological sample containing oocytes, wherein the OCT is operably coupled to an artificial intelligence / machine learning system (AI / ML system) and an imaging system, and the imaging system includes a camera system and an illumination system. The OCT, AI / ML system and imaging system can be used to create at least one three-dimensional image of oocytes. The three-dimensional image can be analyzed using the AI / ML system, and the analysis includes detecting the presence or absence of polar bodies based at least partially on a planar view of the oocyte.

[0109] In some embodiments, the oocyte may be a plurality of oocytes or an embryo.

[0110] In some embodiments, the embryo can consist of multiple embryos.

[0111] In some embodiments, the OCT can be a polarization-sensitive OCT.

[0112] In some embodiments, the OCT can be positioned above the biological sample.

[0113] In some embodiments, the OCT can be positioned below the biological sample.

[0114] In some embodiments, the OCT can be positioned on the side of the biological sample.

[0115] In some embodiments, the OCT can be moved to multiple locations on a biological sample and generate images from those locations.

[0116] In some embodiments, the lighting system may include polarized lighting.

[0117] In some embodiments, a three-dimensional image can be a fusion of multiple images combined with the three-dimensional image.

[0118] In some embodiments, the image fusion body may include image components from OCT and at least one other microscopy system.

[0119] In some embodiments, the 3D image may include simulated virtual elements.

[0120] In some embodiments, the AI / ML system and imaging system can automatically determine the presence or absence of meiotic spindles based at least partially on a three-dimensional image.

[0121] In some embodiments, the AI / ML system and imaging system can automatically evaluate the meiotic spindle to form a predictive algorithm and determine the probability of whether the oocyte will mature sufficiently.

[0122] In some embodiments, the AI / ML system and imaging system can automatically evaluate the meiotic spindle to form a predictive algorithm to determine the probability that the oocyte will mature in further incubation.

[0123] In some embodiments, the AI / ML system and imaging system can automatically determine the best arrangement of oocytes for use during injection.

[0124] In some embodiments, the AI / ML system and imaging system can automatically evaluate membrane integrity.

[0125] In some embodiments, a method for oocyte identification and maturity assessment based on automated artificial intelligence using optical coherence tomography (OCT) is provided, comprising positioning an OCT imaging system head on the side of a biological sample containing oocytes, wherein the OCT is operably coupled to an artificial intelligence / machine learning system (AI / ML system) and an imaging system, and the imaging system includes a camera system and an illumination system. Polarized illumination can be used and directed onto the biological sample containing oocytes. Using the OCT, AI / ML system and imaging system, at least one three-dimensional image of the oocytes can be created. The three-dimensional image can be analyzed using the AI / ML system, and the analysis includes detecting the presence or absence of a meiotic spindle based at least partially on the three-dimensional image, and the maturity assessment of the oocytes can be performed at least partially based on the detection of the presence or absence of a meiotic spindle.

[0126] In some embodiments, the 3D image may include simulated virtual elements.

[0127] In some embodiments, a method for automated artificial intelligence-based IVF microtool control, comprising receiving commands in a control device associated with an artificial intelligence / machine learning system (AI / ML system) and an imaging system, which is at least partially used to operate robotic components of an in vitro fertilization (IVF) module. The robotic system can retrieve at least one IVF microtool assembly (MA) from a tool stock location using a first robotic mechanism of the robotic system, at least partially on command. The robotic system can retrieve at least one IVF receptor using a second robotic mechanism of the robotic system, at least partially on command. The robotic system can position the IVF receptor on a stage within the IVF module, at least partially on command. The first robotic mechanism can be used to position at least one MA in proximity to the IVF receptor, at least partially on command, the positioning in accordance with a planned IVF procedure, its operation being at least partially controlled by the AI / ML system and the imaging system, and can verify the correct positioning of at least one MA in proximity to the IVF receptor within the IVF module, and can send warnings to verify the correct positioning.

[0128] In some embodiments, the first robot mechanism and the second robot mechanism can be a single robot mechanism.

[0129] In some embodiments, the first robot mechanism can be a plurality of robot mechanisms.

[0130] In some embodiments, the second robot mechanism can be a plurality of robot mechanisms.

[0131] In some embodiments, the retrieval of IVF receptors may be from a transport vehicle used to hold or move the IVF receptors.

[0132] In some embodiments, the transport mechanism can be a belt conveyor.

[0133] In some embodiments, the transport mechanism can be a rail conveyor.

[0134] In some embodiments, the transporter may include at least one form capable of holding IVF receptors.

[0135] In some embodiments, the transport system may include a temperature control plate located near the IVF receptor to maintain a predetermined temperature for the biological sample.

[0136] In some embodiments, the imaging system may include a microscope system, a camera system, and a lighting system.

[0137] In some embodiments, the inventory locations on the stage can be multiple locations.

[0138] In some embodiments, the IVF receptor can be a dish, a tube, a microtool, or an ICSI needle.

[0139] In some embodiments, a method for automated embryo manipulation based on artificial intelligence comprises positioning an in vitro fertilization (IVF) receptor containing an embryo on a transport mechanism within an oocyte preparation module using a microtool assembly (MA) controlled by a first robotic mechanism, the transport mechanism comprising at least one form capable of holding the IVF receptor and a temperature control plate adjacent to the IVF receptor to maintain a predetermined temperature for the IVF receptor and embryo, the positioning being performed at least partially using an artificial intelligence / machine learning system (AI / ML system) and an imaging system, the imaging system comprising a microscope system, a camera system and an illumination system. The IVF receptor can be transported from the oocyte preparation module to an incubation module using the transport mechanism, the transport of the IVF receptor being performed to a predetermined location within the incubator using the AI / ML system and the imaging system. IVF receptors can be transported from the incubation module to a position proximal to the microscopy stage after a predetermined period, the position of which is determined at least partially by the AI / ML system and imaging system, allowing for evaluation of the embryo's developmental stage, and the embryo can be transported to the IVF platform position at least partially based on maturity evaluation.

[0140] In some embodiments, the evaluation of the embryonic developmental stage may include imaging to detect the presence of polar bodies.

[0141] In some embodiments, embryonic developmental stage assessment can be performed on multiple embryos within a cohort.

[0142] In some embodiments, a cohort can be defined partially on the basis of embryos that share a common IVF receptor.

[0143] In some embodiments, the AI / ML system can generate a maturity score based at least partially on an assessment of the developmental stage.

[0144] In some embodiments of this disclosure, the semen preparation module may include a robotic system for semen preparation, collection, and transfer to an ICSI dish. In some embodiments of this disclosure, the semen preparation module may include AI for automatic detection, identification, and classification. In some embodiments of this disclosure, the semen preparation module may include AI for automatic measurement and testing. In some embodiments of this disclosure, the semen preparation module may include AI for optimization. In some embodiments of this disclosure, the semen preparation module may include AI for prediction. In some embodiments of this disclosure, the semen preparation module may include AI for selection / grading. In some embodiments of this disclosure, the semen preparation module may include AI for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the semen preparation module may include AI for system configuration and control. In some embodiments of this disclosure, the semen preparation module may include fully autonomous AI. In some embodiments of this disclosure, the semen preparation module may include optical, imaging, and machine vision components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include robotic processing systems and processes. In some embodiments of this disclosure, the semen preparation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include semen preparation components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include laser components, systems, and processes.

[0145] In some embodiments of this disclosure, the oocyte preparation module may include oocyte retrieval components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include AI for automated detection, identification, and classification. In some embodiments of this disclosure, the oocyte preparation module may include AI for automated measurement and testing. In some embodiments of this disclosure, the oocyte preparation module may include AI for optimization. In some embodiments of this disclosure, the oocyte preparation module may include AI for prediction. In some embodiments of this disclosure, the oocyte preparation module may include AI for selection / grading. In some embodiments of this disclosure, the oocyte preparation module may include AI for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the oocyte preparation module may include AI for system configuration and control. In some embodiments of this disclosure, the oocyte preparation module may include fully autonomous AI. In some embodiments of this disclosure, the oocyte preparation module may include optical, imaging, and machine vision components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include robotic processing systems and processes. In some embodiments of this disclosure, the oocyte preparation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include enzymatic oocyte detachment processes, systems, and components. In some embodiments of this disclosure, the oocyte preparation module may include advanced microscopy systems and components.

[0146] In some embodiments of this disclosure, the fertilization module may include fertilization. In some embodiments of this disclosure, the fertilization module may include fertilization components, systems, and processes. In some embodiments of this disclosure, the fertilization module may include AI for automated detection, identification, and classification. In some embodiments of this disclosure, the fertilization module may include AI for automated measurement and testing. In some embodiments of this disclosure, the fertilization module may include AI for optimization. In some embodiments of this disclosure, the fertilization module may include AI for prediction. In some embodiments of this disclosure, the fertilization module may include AI for selection / rating. In some embodiments of this disclosure, the fertilization module may include AI for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the fertilization module may include AI for system configuration and control. In some embodiments of this disclosure, the fertilization module may include fully autonomous AI. In some embodiments of this disclosure, the fertilization module may include optical, imaging, and machine vision components, systems, and processes. In some embodiments of this disclosure, the fertilization module may include robotic processing systems and processes. In some embodiments of this disclosure, the fertilization module may include sensor components, systems, and processes. In some embodiments of this disclosure, the fertilization module may include specimen management components, systems, and processes.

[0147] In some embodiments of this disclosure, the incubation module may include incubation components, systems, and processes. In some embodiments of this disclosure, the incubation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the incubation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the incubation module may include AI for automated measurement and testing. In some embodiments of this disclosure, the incubation module may include AI for optimization. In some embodiments of this disclosure, the incubation module may include AI for prediction. In some embodiments of this disclosure, the incubation module may include AI for selection / rating. In some embodiments of this disclosure, the incubation module may include AI for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the incubation module may include AI for system configuration and control. In some embodiments of this disclosure, the incubation module may include fully autonomous AI. In some embodiments of this disclosure, the incubation module may include robotic processing systems and processes. In some embodiments of this disclosure, the incubation module may include an advanced microscope system and its components.

[0148] In some embodiments of this disclosure, the vitrification and cryopreservation module may include cryopreservation components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for automated measurement and testing. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for optimization. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for system configuration and control. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the vitrification and cryopreservation module may include fully autonomous AI. In some embodiments of this disclosure, the vitrification and cryopreservation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include robotic processing systems and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include storage components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for automated measurement and testing. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for optimization. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for system configuration and control. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the vitrification and cryopreservation module may include fully autonomous AI. In some embodiments of this disclosure, the vitrification and cryopreservation module may include sensor components, systems, and processes.In some embodiments of this disclosure, the vitrification and cryopreservation module may include a robotic processing system and process.

[0149] This disclosure will be better understood from the detailed description and accompanying drawings. [Brief explanation of the drawing]

[0150] [Figure 1] This shows a left side view of the semen preparation module of the IVF / ICSI platform. [Figure 2] This shows a partial assembly diagram of the semen preparation module for the IVF / ICSI platform. [Figure 3] This shows a right-side view of the semen preparation module of the IVF / ICSI platform. [Figure 4] This shows a top view of the semen preparation module of the IVF / ICSI platform. [Figure 5] This shows a front view of the semen preparation module of the IVF / ICSI platform. [Figure 6] This shows a bottom view of the semen preparation module of the IVF / ICSI platform. [Figure 7] This image shows a horizontally configured microscope positioned near a rectangular cuvette containing a biological sample. [Figure 8] This is an example to illustrate images obtained using a horizontally configured microscope placed near a rectangular cuvette containing a biological sample. [Figure 9] This shows the sample screen for sperm tracking analysis. [Figure 10A] This shows a microscope that includes a combination of a CMOS camera sensor and a microscope objective lens. [Figure 10B] This shows the assembled microscope as it is used as an inverted microscope within the IVF / ICSI platform robotic mechanism. [Figure 10C] This shows the appearance of the assembled microscope when used as an inverted microscope within the IVF / ICSI platform robotic mechanism. [Figure 11A]This is a simplified diagram of the components of an inverted microscope assembly. [Figure 11B] This is a simplified diagram of the components of an inverted microscope assembly. [Figure 11C] This is a simplified diagram of the components of an inverted microscope assembly. [Figure 11D] This is a simplified diagram of the components of an inverted microscope assembly. [Figure 11E] This is a simplified diagram of the components of an inverted microscope assembly. [Figure 11F] This is a simplified diagram of the components of an inverted microscope assembly. [Figure 12] The image shows a sample taken with an inverted microscope. [Figure 13A] This shows a microscope light source including an adjustable Lumencor® lamp, which is compatible with optical fibers and features a plano-convex lens. [Figure 13B] A detailed diagram of the microscope light source is shown. [Figure 13C] This shows an exemplary mechanical diagram of this light. [Figure 14A] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14B] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14C] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14D] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14E] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14F] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14G] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 14H] This image system uses a rotating stage mechanically attached to a microscope to partially measure the linear polarization characteristics of light transmitted through a sample. [Figure 15] This shows the linear axis for tool head movement of the pipetting device on the IVF / ICSI platform. [Figure 16] This shows the assembly of components within the pipetting device for tool head movement in the X and Y planes. [Figure 17] This shows pipettes used in IVF / ICSI platform pipetting devices that can generate positive or negative pressure. [Figure 18] The temperature control panel is shown. [Figure 19A] This shows a semen cup holder that can be installed on and / or directly adjacent to a temperature control plate. [Figure 19B] This shows a semen cup holder that can be installed on and / or directly adjacent to a temperature control plate. [Figure 20A] This shows a robotic tilting mechanism capable of holding a specimen or other material. [Figure 20B] This shows a robotic tilting mechanism capable of holding a specimen or other material. [Figure 20C] This shows a robotic tilting mechanism capable of holding a specimen or other material. [Figure 21] This shows a holder for an intracytoplasmic sperm injection (ISCI) dish. [Figure 22] This shows a pipette tip with a custom shape designed to facilitate the sperm "swim-up" protocol. [Figure 23] This shows a pipette tip with a custom shape designed to facilitate the sperm "swim-up" protocol. [Figure 24] This shows a tube holder designed to hold a tube with a conical base. [Figure 25] This shows a tube holder designed to hold a tube with a rounded bottom. [Figure 26] This shows a pipette attachment mechanism that can be attached to a pipette mechanism. [Figure 27] This shows a pipette tip holder and a pipette discarder for automatically discarding pipette tips within the semen preparation module. [Figure 28] This shows the mechanism for discarding the pipette tip. [Figure 29] This shows an angle-controlled holder for clinical specimens. [Figure 30] This shows an angle-controlling holder for clinical use. [Figure 31] This shows an angle-controlling holder for clinical use. [Figure 32A] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32B] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32C] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32D] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32E] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32F] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32G] This shows a robotic pipette for an IVF / ICSI platform. [Figure 32H] This shows a robotic pipette for an IVF / ICSI platform. [Figure 33A] An example of a cartridge for holding the tip of a pipette is shown. [Figure 33B] An example of a cartridge for holding the tip of a pipette is shown. [Figure 33C] An example of a cartridge for holding the tip of a pipette is shown. [Figure 34A]This shows a robotic tilting mechanism that can be used in a swim-up tube as part of a semen preparation module. [Figure 34B] This shows a robotic tilting mechanism that can be used in a swim-up tube as part of a semen preparation module. [Figure 34C] This shows a robotic tilting mechanism that can be used in a swim-up tube as part of a semen preparation module. [Figure 35A] This shows an automated drainage mechanism that can be used with the IVF / ICSI platform. [Figure 35B] This shows an automated drainage mechanism that can be used with the IVF / ICSI platform. [Figure 36] An exemplary form of a dish holder, which can be used by an oocyte preparation module to hold a sample dish, receptor, or container, is shown. [Figure 37] An exemplary form of a dish holder, which can be used by an oocyte preparation module to hold a sample dish, receptor, or container, is shown. [Figure 38] This shows an adapter for using a single-cell holder together with a standard needle holder within the oocyte preparation module. [Figure 39] This shows an ISCI dish adapter that is connected to an aluminum extruded part and can be used by an egg preparation module. [Figure 40] It exhibits a mechanical barrier that interacts with organic compounds. [Figure 41] This document illustrates an exemplary embodiment of an adapter for using a ring light with a standard needle holder that can be used within a fertilization module. [Figure 42] This shows an adapter for using an XY positioning system with a microscope neck that can be used within the insemination module. [Figure 43] This shows a mount for a microscope-compatible cell sorter and / or cell harvester that can be used within the fertilization module. [Figure 44]This shows a device that can be used to rotate one or more knobs of a fluid aspirator that can be used within a fertilization module. [Figure 45] This shows a device that can be used to rotate one or more knobs of a fluid aspirator that can be used within a fertilization module. [Figure 46] This shows a device that can be used to rotate one or more knobs of a fluid aspirator that can be used within a fertilization module. [Figure 47] This shows a device that can be used to rotate one or more knobs of a fluid aspirator that can be used within a fertilization module. [Figure 48] This shows a device that can be used to rotate one or more knobs of a fluid aspirator that can be used within a fertilization module. [Figure 49] This shows an adapter that can be used to connect a linear axis to an XY micropositioning device that can be used within a fertilization module. [Figure 50] An exemplary image frame from a visual servo is shown to position the sperm tail within the laser path. [Figure 51] An exemplary system control diagram is shown. [Figure 52] A simplified system control diagram for laser sperm removal is shown. [Figure 53] A simplified system control diagram for precise needle positioning is shown. [Figure 54] A simplified system control diagram for precise positioning of the retainer is shown. [Figure 55] A simplified system control diagram for the precise positioning of sperm inside the needle is shown. [Figure 56] A simplified system control diagram for a typical cell aspiration configuration is shown. [Figure 57] A simplified diagram of a model-based adaptive control system is shown. [Figure 58]This document presents a simplified workflow for locating sperm droplets and oocytes, which can be used with IVF / ICSI platforms. [Figure 59] This shows an exemplary configuration of an inverted microscope assembly that can be used with an IVF / ICSI platform. [Figure 60] This document shows exemplary assemblies for lenses, lens adapters, cameras, and stage assemblies that can be used with IVF / ICSI platforms. [Figure 61] The IVF / ICSI platform demonstrates an exemplary robotic sperm preparation 6100, which can be performed by pipetting single sperm into groups using a semen / culture mixing and sperm tracking system, and diluting the seminal plasma using a droplet washing system. [Figure 62] The image shows a collection tube 6202 filled with biopsy material from the testis or epididymis aspirate. [Figure 63] This demonstrates robotic sperm preparation without swim-up, centrifugation, or microfluidics, using a horizontal swim-out method. [Figure 64] This shows an adapter that allows a handheld microscope to be connected to a metal extruded product. [Figure 65] This shows an adapter for using an Eppendorf fine adjustment device with metal extrusions. [Figure 66] This shows an adapter for using a rod-shaped light source with a metal extruded part. [Figure 67] This describes a device equipped with suction capability, independent Z-axis control, and a variable volume pipette, with the ability to move visually, XYZ, and fine-tune the device. [Figure 68] This describes a device that can be attached to the neck of a microscope and has independent Z-axis movement and pipetting capabilities. [Figure 69]This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 70] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 71] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 72] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 73] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 74]This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 75] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 76] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 77] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 78] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 79]This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 80] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 81] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 82] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 83] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 84]This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 85] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 86] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 87] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 88] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 89]This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 90] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 91] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 92] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 93] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 94]This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 95] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 96] This is a schematic diagram of an embodiment of a neural network system that can be connected to, integrated with, and made accessible to a platform for enabling intelligent processing, including expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, pattern recognition systems, classification of one or more parameters, characteristics or phenomena, support for autonomous control, and other purposes. [Figure 97] These are electron microscope images of unstandardized images used as input for image standardization-based methods for classifying human embryonic cells. [Figure 98] This is an electron microscope image of a second, unstandardized image used as input for an image standardization-based method for classifying human embryonic cells. [Figure 99] This is an electron microscope image of a third unstandardized image used as input for an image standardization-based method for classifying human embryonic cells. [Figure 100] These are electron microscope images of the first images after preprocessing and / or adjustment, which have been preprocessed with standardization parameters used in image standardization-based methods for classifying human embryonic cells. [Figure 101]These are electron microscope images of a second set of images that have been preprocessed and / or adjusted, and which have been preprocessed with standardization parameters used in image standardization-based methods for classifying human embryonic cells. [Figure 102] These are electron microscope images of a third image that has been preprocessed and / or adjusted, and which has been preprocessed with standardization parameters used in image standardization-based methods for classifying human embryonic cells. [Figure 103] This is a diagram of the elements constituting Example 1 of a method based on image adjustment and preprocessing for classifying human embryonic cells. [Figure 104] This demonstrates image segmentation using registered embryo images. [Figure 105] This diagram illustrates a system for real-time automated quantitative evaluation, assessment, and / or classification of individual sperm. [Figure 106] This is a diagram of equipment (camera and microscope) that generates images of a system for real-time evaluation, assessment, and / or automated quantitative classification of individual sperm. [Figure 107] This is a diagram illustrating the semantic division of sperm in a video sequence. [Figure 108] This diagram illustrates the semantic partitioning of sperm in a video sequence using neural network structures. [Figure 109] This is a diagram illustrating the process of verifying the correspondence of sperm identity within a continuous frame. [Figure 110] This diagram shows examples of inputs and outputs for the procedures described herein. [Figure 111] This diagram shows examples of inputs and outputs for the procedures described herein. [Modes for carrying out the invention]

[0151] In some embodiments of the present invention, the IVF / ICSI platform described herein may include, but is not limited to, the automation of conventional manual laboratory work between robotic systems used for assisted reproductive technology (between robots) and the automation of conventional manual laboratory work to prepare for specific robotic procedures (in-robot), 1) diagnostic semen analysis such as computer-assisted sperm analysis, 2) continuous embryo culture including the use of robotic manipulation of dishes combined with time-series microscopy techniques with manual or automated embryogenesis annotation, 3) cryopreservation automation including an automated cryopreservation process in clinical IVF that enables precise sample location monitoring and continuous environmental monitoring, and 4) the use of fine-tuning devices including robotic systems for single-cell surgery, providing fine and coarse-control movement tuners with some digital control for precise tool manipulation.

[0152] In some embodiments, the IVF / ICSI platforms described herein improve existing methods and systems by integrating conventional IVF laboratory processes, including but not limited to automated semen analysis, sperm preparation, petri dish preparation, oocyte retrieval, oocyte vitrification, oocyte detachment, time-course incubation, ICSI, embryo selection, embryo vitrification, robotic immersion in liquid nitrogen, embryo transfer, cryopreservation, and other IVF systems, methods, processes, and procedures, into a conveyor-type robotic line (linear, etc.) and by automating in-patient or inter-patient pipetting and tool configuration, dish preparation, or tracking of disposable supplies and biological materials. In some embodiments, transport systems used by the IVF / ICSI platform to transport tools, equipment, biological materials, human samples, waste, and other facilities, components, materials and / or objects from a first location or position to a second location or position include, but are not limited to, conveyor belts, rail transport systems, continuous robotic transport, roller conveyors, chain conveyors, gravity conveyors, overhead conveyors, flexible conveyors, pneumatic conveyors, auger conveyors, screw conveyors, vacuum conveyors, vibratory conveyors, or any other type of transport system.

[0153] In some embodiments, the IVF / ICSI platform may include an inter-robot IVF system protocol that includes a process for coordinating robotic elements across different IVF procedures, specifically addressing the interconnection of robotic stages within the IVF / ICSI platform, thereby reducing human intervention, minimizing costs, improving operating speed, and ensuring the safe and efficient transfer of samples between different robotic modules of the IVF / ICSI platform.

[0154] In some embodiments, the IVF / ICSI platform may include a comprehensive documentation system for recording, monitoring, and reporting the performance of the robot-to-IVF system, including parameters related to sample transfer, system efficiency, and any events for analysis and continuous improvement. The IVF / ICSI platform may be adaptable to changes in the IVF laboratory configuration and may be scalable to accommodate future expansions or modifications.

[0155] In some embodiments, the IVF / ICSI platform may include in-robot systems and procedures for configuring an in-robot IVF system that incorporates microscopic and non-microscopic robotic elements involved in specific procedures. Such in-robot systems and procedures can standardize the configuration for each patient and ensure the incorporation of both disposable and non-disposable components into each robotic platform. An exemplary in-robot IVF system procedure can verify the availability of components required for a specific patient procedure, and the procedure can ensure that both disposable and non-disposable supplies are in stock and within a designated sterile environment. Sterile and non-sterile components can be arranged on the robotic platform according to a standardized layout, and in part, the flexibility of the placement of individual components allows for adaptation to variations in patient or procedure requirements and facilitates adjustment of the configuration. Sterilization protocols can be performed by the system when handling and placing sterile components on the robotic platform, which can also periodically assess and maintain the integrity of the sterile barrier throughout the procedure. The system can document the configuration of the in-robot IVF system for each patient procedure, including details on the placement of disposable and non-disposable components, enabling comprehensive record-keeping and potential future optimization.

[0156] In some embodiments, the inter- and intra-robot systems of the IVF / ICSI platform may include a comprehensive software system for coordinating and managing the IVF process and integrated laboratory robots. This system can track samples, monitor environmental conditions, monitor and direct the robotic system, control the timing of all procedures, detect failures or inefficiencies, and report to the electronic medical record (EMR) to enable patient scheduling and patient instruction. The software system can monitor the safety aspects of the robotic system, including maintenance and service requirements.

[0157] In some embodiments of the present invention, the IVF / ICSI platform described herein can receive semen samples from a robotic device such as an operating robot, a liquid processing machine, or some other type of automated process, rather than from a person manually placing the semen sample into the semen preparation module of the IVF / ICSI platform. Figure 1 shows a left side view of the semen preparation module 100 of the IVF / ICSI platform. Figure 2 shows a partially assembled view of the semen preparation module 200 of the IVF / ICSI platform. Figure 3 shows a right side view of the semen preparation module 300 of the IVF / ICSI platform. Figure 4 shows a top view of the semen preparation module 400 of the IVF / ICSI platform. Figure 5 shows a front view of the semen preparation module 500 of the IVF / ICSI platform. Figure 6 shows a bottom view of the semen preparation module 600 of the IVF / ICSI platform.

[0158] In some embodiments, the semen preparation module may be fully automated, using robotic mechanisms for processing and moving materials, including biological specimens, and using automated robotic pipetting operations. It may also be connected to the network infrastructure described herein for remote control of the semen preparation module's operation as a component of a fully automated robotic IVF / ICSI platform. Semen is a complex bodily fluid consisting primarily of a mixture of sperm and bodily fluids produced by the seminal vesicles and prostate gland. This bodily fluid, or seminal plasma, performs several functions, including the transport and protection of spermatocytes. To remain functional and maintain motility and viability, sperm can be separated from the seminal plasma within approximately one hour of production.

[0159] In some embodiments, the automated robotic pipetting systems and methods described herein are fully automated and use robotic mechanisms for handling and moving materials, including biological specimens, and can operate within a connected network infrastructure for remotely controlling the work of single-cell based methods, processes, protocols, etc., including, but not limited to, other robotic pipetting systems and methods, including, but not limited to, methods for single-cell isolation, methods for rapid isolation of live cells into a container (e.g., a PCR tube), methods for isolating live single cells in volumes of less than a microliter (e.g., suitable for single-cell PCR analysis and RNA sequencing), robotic pipetting ("robot pipetter") methods using a single-cell pipette (SCP) consisting of, for example, an air-displacement pipette and an SCP-tip, its application in single-cell isolation for sequencing (e.g., RNA sequencing), and other robotic pipetting systems and methods.

[0160] In some embodiments of the present invention, the term “selection” may refer to the process of selecting specific spermatocytes from a group using certain criteria (e.g., sperm morphology, sperm motility, etc.). For example, the semen preparation module described herein may select individual sperm based, for example, on the pattern of cell motility. The terms “preparation” or “separation” as used herein may be used in relation to semen and in relation to the processes that sperm undergo before they can be used for IVF. Thus, there is an element of selection in sperm preparation (selecting a population of forward-motile sperm), and the term “preparation” as used herein to disclose the semen preparation module fully encompasses the processes of selection, preparation and separation, as well as other functionalities, as described herein.

[0161] In some embodiments, in the context of semen and sperm preparation, the term “automation” as used herein may refer to processes and protocols that include robotic mechanisms, robotic mechanism assistance, and / or the use of robotic mechanisms, including robotic mechanisms that operate without human intervention for the duration of the process or protocol. Accordingly, processes, methods, and protocols that are carried out in semen preparation modules that normally require an operator may be carried out by a robotic system as described herein.

[0162] In some embodiments, for in vitro oocyte fertilization or in vitro artificial uterine delivery of sperm, semen is processed to remove seminal plasma, cells and cell-free debris, and immotile spermatids, and to collect a predominantly motile sperm fraction free of seminal plasma. Seminal plasma is toxic to oocytes and embryos. Table 1 lists the common processes involved in semen preparation. Conventional manual methods, excluding microfluidic separation (for example, for intracytoplasmic sperm injection (ICSI)), require centrifugation of a mixture of semen and processing medium. In standard IVF fertilization, microfluidic separation requires centrifugation to remove trace amounts of seminal plasma.

[0163] Table 1 below lists the processes involved in conventional manual semen preparation via three methods (swim-up, density gradient, and microfluidics) and illustrates the role of the operator in conventional non-robot procedures.

[0164] [Table 1]

[0165] In some embodiments, the preparation for non-diagnostic semen evaluation for fertilization involves measuring volume (mL), viscosity (normal / high), and sperm concentration (×10). 6 This may include measuring sperm motility (per mL), as well as forward and non-forward motility (%). The time required for liquefaction can also be evaluated. Liquefaction is temperature-dependent and should be completed within 15-30 minutes after preparation. Post-preparation evaluation may include sperm concentration and forward motility.

[0166] In some embodiments, a robot, fluid handler, or other type of automated process can be controlled within and / or remotely from the IVF / ICSI platform using computer code, commands, instructions, etc., received from a system associated with the IVF / ICSI platform but physically remote. For example, an operator such as a clinical embryologist and / or male pathologist located away from the IVF / ICSI platform can use a user interface to select a sperm sample, confirm that it is the sperm sample intended for preparation, and instruct a robot, fluid handler, or other type of automated process to move the sperm sample from a first location (e.g., a secure storage location) to the IVF / ICSI platform for preparation.

[0167] In some embodiments, the semen sample can be received in a semen preparation module within a container, for example, a container made of plastic or some other material. Processing of the semen sample by the semen preparation module can be carried out in the same container or in a separate container.

[0168] In some embodiments, the semen sample may be placed in a semen preparation module in a container, vessel, dish, or other holder that comes into contact with a plate that can be moved to optimally position the semen sample, for example, to optimize computer imaging of the sample, sperm detection, or some other operation.

[0169] A semen sample can be placed on, near, above, or next to a plate to which temperature control can be applied (e.g., heating). In one example, the plate can be heated to a sustained temperature of 37 degrees Celsius (or some other temperature target) during the pipetting process. In another example, the plate can have a temperature that is adjusted throughout the entire duration of the pipetting process, for example, based on data regarding the detected state or the pipetting process. This temperature adjustment can be performed automatically, for example, using an AI / ML process as described herein. Alternatively, the semen sample can be placed on, near, above, or next to a plate to which temperature control is not applied.

[0170] In some embodiments, the semen preparation module can receive a sample and automatically position it for examination using a microscope, computer or machine vision or any other imaging device for sperm identification, discovery, analysis, and evaluation. In one exemplary embodiment, the sample can be examined within the semen preparation module using one or more flexible or fixed-position microscope objectives, an inverted microscope, a digital microscope or any other type of microscope.

[0171] In some embodiments, the IVF platform may include imaging systems and methods for tracking sperm motility as sperm swim against gravity, and which can be directly measured for sperm statistics, including concentration, swimming direction, position, and velocity. In contrast to conventional methods in which sperm are observed, for example, through a counting plate on a microscope, this method can allow for direct observation of sperm behavior in a test tube. In some embodiments, this method may include the use of a horizontally configured microscope positioned near a rectangular cuvette containing the sample (700) (see Figure 7). The sample may include two liquid phases: a semen sample at the bottom and a cell treatment medium at the top (as used herein, “culture” and “multiple cultures” may include, but are not limited to, multipurpose treatment medium (MHM), culture medium solution, or other culture or treatment media). A lamp designed for fluorescence experiments may emit light that can be focused onto the treatment medium layer to create an illumination field. Because the two layers are in contact, sperm can swim from the bottom to the top layer, consistent with accepted recommendations for sperm preparation in IVF procedures. By positioning the microscope close to the cuvette, images can be captured, for example, on a CMOS sensor, allowing observation of sperm swimming in the processing medium (as used herein, “image,” “multiple images,” and “optics” include, but are not limited to, images taken from real life without modification, enhanced images, hybrid images including images combining elements of real life images, simulated images, virtual reality images, augmented reality images, graphic elements, moving images, three-dimensional images, two-dimensional images, still images, images over time, video images, or any other type of image and / or visual depiction). To facilitate this, the sample can be mounted on an X, Y, Z fine-tuning device. Figure 8 shows an example for illustrating images obtained through this configuration. Figure 8 shows an example for illustrating images obtained using a horizontally configured microscope placed close to a rectangular cuvette containing a biological sample (800). Figure 9 shows a screen for sperm tracking analysis using custom software (900).

[0172] In some embodiments, the IVF platform may include a specially fabricated microscope for observing various types of cells associated with the IVF procedure and for producing microscopic images very similar to those produced by commercially available stereomicroscopes. An optimization process may be used to ensure that the size, shape, and weight are configured to integrate with a robotic system without compromising image quality. Primary cell types intended for observation with this microscope include sperm, oocytes, and cumulus cells. In some embodiments, one central component of the microscope is the objective lens. Regardless of its size or unique functionality, the essential element of this instrument is a set of lenses that facilitate the formation of a magnified image of the object being examined. The resulting image can be observed directly on the human retina or via its electronic counterpart, for example, with a CMOS camera. As shown in Figure 10A, the microscope may include a combination of a CMOS camera sensor (e.g., IMX477) and a microscope objective lens (1000). In this configuration, the microscope objective lens can perform the function of a camera lens.

[0173] In some embodiments, a focusing lens may be positioned above the sample to improve image contrast. The function of the focusing lens includes focusing light onto the sample while imparting specific properties to the light, such as polarization, phase, and structural characteristics. Images captured by the CMOS sensor can be utilized through various imaging systems, webcams, imaging, and optical applications (including mechanical and computer vision as described herein, collectively referred to as “optical systems,” “imaging systems,” “imaging systems,” etc.). Furthermore, the resulting images can be made available through programming languages ​​such as Python®, MATLAB®, and UNITY via a USB port or other connectivity means that can facilitate automation.

[0174] Figures 10B and 10C show the assembled microscope (1002) and its appearance when used as an inverted microscope within the IVF / ICSI platform robot mechanism (1004), respectively. Figures 11A to 11F (1100 to 1110) show simplified component diagrams of the inverted microscope assembly. Figure 12 shows two examples of images captured by the inverted microscope (1200).

[0175] In some embodiments, high-quality imaging often relies on effective illumination. Even state-of-the-art microscope models may be insufficient to produce optimal results without a good light source. Achieving a uniform spot and quality contrast for imaging biological samples is crucial for obtaining useful imaging results. In some embodiments, the IVF / ICSI platform may include a microscope light source capable of producing high-quality images, which can generate a uniform light spot with an intensity favorable for producing balanced contrast within the image. As shown in Figure 13A, the microscope light source may include, for example, an adjustable Lumencor lamp fitted to an optical fiber and accompanied by a plano-convex lens with dimensions such as a diameter of 25 mm and a focal length of 100 mm (1300). This optical arrangement can produce parallel light characterized by a spot diameter of 40 mm. Figure 13B shows a detailed diagram of the microscope light source 1302. This configuration may incorporate a linear polarizer and a neutral density (ND5) filter for precise control of light intensity to prevent sensor saturation. To diffuse and homogenize the emitted light and improve the overall quality of the illumination, frosted glass can be strategically employed. An exemplary mechanical diagram of this light source is shown in Figure 13C (1304).

[0176] In some embodiments, IVF / ICSI platform imaging systems and methods can partially measure the linear polarization characteristics of light transmitted through a sample using a rotating stage mechanically coupled to a microscope. The microscope can be equipped with an analytical polarization function to enable observation of birefringence phenomena in biological samples. This mechanism can consist of a linear polarizer placed between the microscope's objective lens and camera sensor. To fix the polarizer, a set of mechanical components can be installed, as shown in Figures 14A to 14H (1400 to 1414). The rotation function can facilitate the analysis of polarization states, and the polarizer needs to be set to angles such as 0°, 45°, -45°, and 90°. Rotating the polarizer at a constant angular velocity can facilitate the observation of flickering at certain image points, which can indicate changes in optical properties at those locations. Control can be provided, as described herein, for example, by a stepping motor connected to an electronic interface and control software.

[0177] The IVF / ICSI platform described herein may utilize optical coherence tomography (OCT) in the context of the IVF / ICSI platform. OCT is a non-invasive imaging technique that provides high-resolution three-dimensional images of biological tissues. For example, the IVF / ICSI platform can use OCT to collect detailed information about sperm, oocytes, embryos, or other biological materials. The OCT system of the IVF / ICSI platform can be modified to better suit the specific requirements of the process. For example, the head of the OCT system can be moved to the bottom of the sample, similar to an inverted microscope, to better visualize samples in a liquid state. Furthermore, the OCT system can be modified to be controlled by AI / ML and / or custom software, as described herein, enabling more precise automated control of the imaging process. Figure 59 shows an exemplary configuration of an inverted microscope assembly 5900 that can be used with the IVF / ICSI platform. Figure 60 shows an exemplary assembly for a lens, lens adapter, camera, and stage assembly 6000 that can be used with the IVF / ICSI platform.

[0178] In some embodiments, the IVF / ICSI platform may include a computer-assisted sperm analysis system and a semen preparation robot. Before preparation, semen samples can be evaluated by performing a limited semen analysis. After preparation, the same evaluation can be repeated to determine sperm recovery rate and quality. Computer-assisted sperm analysis can be performed automatically by inserting the sample into the computer-assisted sperm analysis robot system. The robotic mechanism of the semen preparation module can control the computer-assisted sperm analysis, move the sample into the computer-assisted sperm analysis system, and issue warnings when certain criteria, such as acceptable sperm concentration or motility, are met or not met. The semen preparation robot may also adapt the preparation protocol based on initial or intermediate computer-assisted sperm analysis findings. Other functions and procedures may include, but are not limited to, the following:

[0179] Prostate-specific antigen (PSA) testing: The development of PSA testing for the final sperm suspension can be considered to confirm the absence of seminal plasma components.

[0180] Swim-up efficiency and intelligent sperm preparation: By integrating a swim-up efficiency robot based on optical evaluation during swim-up, it is possible to develop optimizations for each sample processing. The robot can also adjust sperm preparation protocols and provide detailed reports for each sample.

[0181] Safe storage of prepared samples: Sperm preparations can be stored separately in a "sperm preparation room" (part of the semen preparation robot) equipped with an evidence system to ensure strict separation of the samples.

[0182] Sample Scheduling (AI Optimization): This system can plan sample schedules, optimize procedure flows, and ensure efficiency and minimal manual intervention. The system can determine which samples are needed at specific stages of the IVF process. Multiple simultaneous processing can be considered depending on the size of the laboratory.

[0183] Follicular fluid processing: Automation can include collecting follicular fluid into tubes and automatically draining the supernatant of these fluids into petri dishes of different diameters, eliminating the need for manual intervention. Automation can include a process for automatically spreading follicular fluid into dishes of different diameters, draining the supernatant, or a high-volume pipetting system.

[0184] Disposable Pasteur Pipette Aspiration System (PPSS): A PPSS can be pre-manufactured, pre-assembled, and sterilized, and consists of a Pasteur pipette-like glass tip of a specific length connected to a tube using a gastrointestinal tube-like tubing similar to that of Nutricat. A PPSS can be connected to a pipette such as a Sartorius or equivalent pipette. A stepper system can be applied when the volume limit is sufficiently small; otherwise, an adjacent control system can be applied.

[0185] Rail Conveyor System: A robotic rail system can be used to transport dishes containing bodily fluids, eggs, sperm preparations, embryos, biopsy samples, and embryos or oocytes for cryopreservation to various robotic systems. It can have a versatile, pre-cut form for accommodating dishes or tubes. The system can move the material at a speed fast enough to minimize exposure to room conditions, and the plate holder can be temperature-controlled to ensure that the dishes are kept between room temperature and 37 degrees Celsius. pH changes can be minimized, requiring gentle but gradual operation of the plate / tube holder conveyor system.

[0186] Dish handling: This system allows for the integrated transport of dishes containing eggs from incubation to the ICSI platform and vice versa, ensuring sample integrity and minimal damage.

[0187] ICSI Platform: The ICSI platform can be equipped with areas for receiving and unloading trays, and loading and unloading procedures can be recorded and verified using an evidence system.

[0188] ICSI Disposable Microtool Assembly (DMA): The DMA system can include a microtool holder system for the precise control of microtools such as syringes, tubes, and ICSI needles, holding pipettes, etc. The assembly can be premanufactured, simplify conventional systems, and adapt to various options for a micromanipulation system with adapters suitable for different micromanipulation device models.

[0189] Robot Arm: The robot arm can ensure accurate alignment, positioning, and optimal tool orientation, for example, as described herein, by using the AI / ML system and the optical system bidirectionally to install the DMA and confirm it by the AI / ML system and the optical system.

[0190] Installation of the ICSI Dish: The ICSI dish with gametes can be placed using the robot arm under the DMA and the microtool tip.

[0191] Alignment and ICSI Procedure: The DMA can be lowered to a predetermined position and the ICSI procedure including the alignment of the tool tip can be started.

[0192] Addition and Removal of Eggs and Sperm during ICSI: This system can enable pipette operation by a third micromanipulation device in case of unexpected events during ICSI.

[0193] Oocyte Confirmation: Before and after the ICSI procedure, mature oocytes can be reevaluated to reconfirm their maturity.

[0194] The platform can autonomously execute 20 - 30 protocols for sperm injection into eggs.

[0195] Processing of Sperm-Injected Eggs: Sperm-injected eggs can be examined for viability, imaged, and placed using a conveyor system into a thermostat robot.

[0196] Thermostat confirmation: The thermostat can confirm the receipt of Petri dishes, monitor embryo development, and predict blastocyst formation based on time-lapse (TL) data.

[0197] Timing of vitrification: The system can predict the optimal timing for vitrification considering the development and expansion of blastocysts. It can support multiple vitrification options and the system can optimize the timing. Blastocysts can be vitrified as a cohort (sequentially) or individually.

[0198] Blastocysts assigned to a vitrification or embryo transfer robot: The optimal development of embryos in a cohort can be confirmed, at least in part, by (a) the TL system being able to confirm optimal development, (b) non-invasive PGT or niPGT being able to provide a euploid indication, or (c) using an AI / ML system aimed at optimizing embryo selection such as ERICA. After embryo selection and scoring are complete, the Petri dish can be transferred to a vitrification robot or one or more selected embryos can be transferred to an embryo transfer robot.

[0199] Treatment of individual embryos: Fast or slow embryos can be individually evaluated to determine the optimal timing of the culture period and vitrification. This may require automated pipetting operations in a separate robotic system aimed at moving eggs and embryos between or within dishes.

[0200] Handling of cryogenic devices: Modified RFID cryogenic vials can be used.

[0201] In some embodiments, terms used herein such as “pipette,” “pipette operation,” “automatic pipetting,” and “robot pipetting” refer to the process of transferring body fluids, cell suspensions, and individual or group cells (oocytes and embryos) and / or single cells from one container to another, using a device capable of generating controlled positive and negative pressure. The process may involve precise aspiration and dispensing, stratification, and / or mixing of body fluids. Pipetting is fundamental to IVF laboratory techniques (e.g., semen preparation, culture medium and dish preparation, oocyte retrieval, ICSI, vitrification, embryo transfer, etc.) and can be performed in a sterile manner without losing cells and without generating bubbles that can be destructive, lead to cell loss, or generate potentially infectious aerosols. Traditionally, pipetting has remained manual in IVF. Manual pipetting requires precision, dexterity, and hand-eye coordination. It may or may not require a microscope, depending on the procedure.

[0202] In some embodiments, the semen preparation module may include a robotic pipette capable of Cartesian movement in the X, Y, and / or Z planes, with independent degrees of freedom of movement between the pipette and the pipetting device to which the pipette is attached. Referring to Figure 15, a linear axis for the movement of the tool head of the pipetting device is shown (1500). Figure 16 shows the assembly of instruments within the pipetting device for the movement of the tool head of the pipetting device in the X and Y planes (1600). As shown in Figure 17, the pipette used in the pipetting device may generate positive or negative pressure (1700).

[0203] In some embodiments, different types of pipettes can be used during semen preparation, and some can be used during other IVF procedures.

[0204] Serological pipettes: Long, slender polystyrene pipettes (typically 1-5 mL) used for measuring seminal volume, preparing density gradients (typically 1-2 mL), and transferring culture media and semen into tubes (typically 1-5 mL). Serological pipettes are often used in conjunction with electric pipettes or "pipette control devices" equipped with filters and capable of operating at different speeds (low and high).

[0205] Pasteur pipettes are made of borosilicate glass, come in various lengths, and are used with rubber pipette valves to transfer smaller amounts of culture medium or sperm suspension. Pasteur pipettes can be "stretched" using a flame to create extremely thin (and long) tips that can handle single cells / embryos.

[0206] Volumetric pipette: A low-density polyethylene pipette molded into a single unit to remove the rubber bulb and reduce the operator's exposure to the sample.

[0207] Eppendorf pipette: A device equipped with a piston and spring-loaded toothed cone, a single pathway, and adjustable volume (in units of 1 to 1000 μL with a specific range), used in conjunction with an Eppendorf tip to aspirate and dispense precise (usually low) volumes.

[0208] In some embodiments, the IVF / ICSI platforms described herein can incorporate pipetting using robotic mechanisms to replace manual, human-operated pipetting. The IVF / ICSI platforms may not require centrifugation during semen preparation for conventional insemination and ICSI. The robot-controlled, highly precise pipetting described herein can significantly reduce both the complexity and cost associated with the microfluidic chamber while eliminating the need for centrifugation. Centrifugation of semen can result in the generation of oxygen-free radicals, which can damage sperm membranes and potentially reduce their viability and fertilization ability. While samples prepared using microfluidic devices can be used in ICSI procedures, the potential for contamination of the final sample in seminal plasma due to design and inaccurate manual pipetting makes these samples unsuitable for conventional insemination procedures, requiring the addition of a "washing" process via centrifugation.

[0209] In some embodiments, the IVF / ICSI platform may include a metal plate maintained at 37°C (e.g., continuously monitored through sensors and digital readouts) (1800) or some other temperature (see Figure 18). The surface area of ​​the plate can be heated / cooled to a specified temperature to avoid fluctuations and uniform heat distribution. The plate may incorporate sensors that allow its temperature to vary (e.g., between room temperature and body temperature). An algorithm may be used to set and maintain the temperature.

[0210] In some embodiments, the IVF / ICSI platform may include a semen cup holder mounted directly on and / or adjacent to the heating plate (1900, 1902) (see Figures 19A and 19B).

[0211] In some embodiments, the IVF / ICSI platform may include a 3D-printed specimen holder. The holder may be made of plastic, metal, or some other material. The holder may allow the bottom of the specimen cup to be in sufficient contact with the heating plate. The holder may be modified to accommodate specimen containers of different sizes and shapes.

[0212] In some embodiments, the IVF / ICSI platform may include a robotic tilting mechanism capable of holding a specimen or other material (2000, 2002, 2004) (see Figures 20A–20C). In some embodiments, the oscillating movement may be organized using software code, optics, AI / ML and other processes as described herein to control the position, timing and speed of the retainer. In some embodiments, the tilting mechanism may incorporate a servomotor (e.g., equipped with a sensor) having feedback capability that can be controlled to define the degree of tilt and the speed of the oscillating movement. The retainer may be held at a specific angle depending on the process being performed.

[0213] In some embodiments, the tilting motion, degree of tilting, speed of tilting, duration of tilting, and other tilting dynamics can be automatically adjusted and optimized by the IVF / ICSI platform to improve sperm processing, enhance the hydrodynamic and mechanical advantages present in human sperm sample processing, and influence the normalization of fluid volume / density and viscosity. In one embodiment, Figure 21 shows a holder for an intracytoplasmic sperm infusion (ISCI) dish (2100). In one embodiment, Figures 22 and 23 show pipette tips with custom shapes to facilitate sperm "swim-up" protocols (2200, 2300).

[0214] In one embodiment, the tube holder can be designed to hold tubes having a conical bottom 2400 (see Figure 24), a round bottom 2500 (see Figure 25), a flat bottom, or any other configuration. A pipette attachment mechanism 2600, as shown in Figure 26, can be attached to the pipette mechanism (2600). As shown in Figure 27, the semen preparation module may include a pipette tip holder 2700 and a pipette discarder 2800 for automatically discarding pipette tips (Figure 28). Tilt of sperm samples may include the use of an angle-controlled holder 2900 for clinical samples (Figure 29), an angle-controlled holder 3000 for clinical tubes (Figure 30), and a tilt mechanism 3100 (Figure 31).

[0215] In some embodiments, the IVF / ICSI platform may include a robotic pipette (3200-3214) (see Figures 32A-32H). In some embodiments, the robotic pipette assembly may include, but is not limited to, 1) the functional components of a manual pipette, consisting of, for example, a plunger, a lubrication spring, and a body / cylinder; 2) a high-precision motor (e.g., a stepping motor) that drives the plunger; the plunger moves along a linear rail / track; and 3) a second, more powerful motor with a second rail that drives components 1 and 2 in the Z direction. Components 1, 2 and 3 may be moved in the X and Y directions with controlled speed and acceleration using proprietary software.

[0216] In some embodiments, the robotic pipette of the IVF / ICSI platform may include a pipette tip cartridge (3300-3304) (see Figures 33A-33C) and may include 3D printed parts (plastic or metal) for accommodating multiple numbers and sizes of pipette tips to handle different capacities, tip brands, and different tube heights.

[0217] In some embodiments, the robotic tilting mechanism described herein can be used in a swim-up tube as part of a semen preparation module (3400 - 3406) (see FIGS. 34A - 34C). The IVF / ICSI platform assembly can include a tube holder, a motor, a bracket that holds the motor to the robotic frame, and a control arm. The tilting movement can be choreographed using software code, optics, an image system, AI / ML, and other processes as described herein to control the position, timing, and speed of the movement of the tube holder, and can incorporate a motor (e.g., equipped with sensors) that has feedback capabilities to control the degree of tilt and the speed of the tilting movement. The position of the holder can be held at a specific angle depending on the process being performed. The holder can incorporate a window or notch to enable visualization of the tube during the procedure. In some embodiments, multiple tube holder types can be incorporated into the IVF / ICSI platform. In one example, two of three tubes can remain stationary / static in a standing position, and the remaining third can be tilted. The holder can be designed to hold tubes of multiple sizes and enable visualization of the contents through the window / notch.

[0218] In some embodiments, the IVF / ICSI platform can include an automatic ejection mechanism (3500, 3502) (see FIGS. 35A and 35B). In one example, the automatic ejection mechanism can include a 3D printed (plastic or metal) part that fits over the top of a manual pipettor body and enables upward movement of the pipettor (facilitated by a z - motor as described herein) to capture the pipettor tip and separate it from the pipettor body, and can include a vertical part with a fork - like appendage.

[0219] In some embodiments, in an exemplary use of the IVF / ICSI platform, the automated robotic pipetting within the semen preparation module can follow the simplified protocol described below. 1. The operator can prepare the semen preparation module robot processing array by placing the tubes in the designated tube holders (3400-3406) (Figures 34A-34C) and the multiple pipette tips in the tip cartridges (Figures 33A-33C). Alternatively, as described herein, this process can be performed by a robot without the use of an operator. 2. A waste bag can be attached to the outside of the frame for disposing of used pipette tips. 3. An operator can place the sample cups in the cup holders 1900 and 1902 (Figures 19A and 19B) on the heating plate 1800 (Figure 18). Alternatively, as described herein, this process can be performed by a robot without the use of an operator. 4. An operator can initiate the process (and calibration routine) in the semen preparation program by issuing one or more commands to the IVF / ICSI platform. Alternatively, as described herein, this process can be performed automatically without the use of an operator. 5. The robotic tilting mechanism (Figures 20A-20C) can tilt the cup to a specified degree and then return the cup to its original upright position (2000, 2002, 2004). This can be repeated several times (to facilitate liquefaction). The tilting can be performed according to a preset operation and / or adjusted in real time according to an ML / AI-based algorithm. For example, during the tilting process, machine vision can detect the properties (e.g., viscosity) of the sample held in the cup held by the tilting mechanism, and based at least in part on these properties, the IVF / ICSI platform can command new operating parameters, including, but not limited to, robotic control settings affecting the tilting angle of the cup, the travel speed during the tilting operation, and the duration of the tilting operation. 6. The robotic pipette (Figures 32A-32H) can move to the position of the pipette tip cartridge (X and Y axes) and / or move downward (Z axis), attach the pipette tip to the pipette, and return to the zero position (upper position) (Z axis) (3200-3214). 7. Once the tip is attached, the robotic pipette moves to the position of the sample cup (X and Y axes), and while the cup is tilted by the robotic tilting mechanism, the pipette / tip can be lowered into the cup (Z axis). 8. A robotic pipette can aspirate a specified volume of semen and discharge that volume back into a sample cup (simulating mixing). 9. The robotic pipette can be lifted from the sample cup (Z-axis) and kept stationary in that position (specified height, zero position, or uppermost position) for a specified time. 10. By repeating the process for a specified time, liquefaction can be achieved. 11. After the specified time has elapsed, the robotic pipette can descend into the sample cup (Z-axis) while the cup is in an inclined position, aspirate a specified volume of semen, and raise the sample cup to a specified height (Z-axis) while the robotic rocker returns to an upright position. 12. The robotic pipette moves precisely above the swim-up tube (X and Y axes), descends to the bottom of the tube (Z axis), and can discharge semen into the tube at a specified rate and acceleration without generating bubbles or aerosols. Process 11 or 12 can be repeated until the desired final volume is obtained. 13. The robotic pipette can move (X and Y axes), stop above the waste bag, and discharge the used pipette tip into the waste bag (using X, Y, and Z motion) (3500, 3502) (Figures 35A and 35B). 14. The robotic pipette can then move to the position of the pipette tip cartridge (X and Y axes), descend (Z axis), and attach a new pipette tip to the pipette. 15. The robotic pipette can move to a position directly above the tube containing the sperm processing / washing medium (X and Y axes), descend into the medium to a specified depth (Z axis), aspirate a specified volume of medium, lift the tube (Z axis), move to the tube containing the semen, descend to a precise position above the semen meniscus in the tube (Z axis), and "layer" the medium on top of the semen at a specified speed and acceleration. Process 15 can be repeated until the desired final volume is obtained. 16. The robotic pipette can then move upward (Z-axis) and exit the tube. Process 13 can be performed to discharge the used pipette tip, and the pipette can remain stationary for a specified time at a designated position next to the waste bag while the swim-up is in progress. 17. After the specified time for swim-up has elapsed, process 14 can be repeated to attach a new pipette tip. 18. The robotic pipette can then move to a position directly above the swim-up tube (X and Y axes), descend to a specified depth in the swim-up layer (Z axis), aspirate a specified volume of cell suspension, move to a position above the final tube (X and Y axes), descend into the tube (Z axis), and discharge the volume at the bottom of the tube. 19. The robotic pipette can then be lifted from the tube, moved to a waste bag, and the used pipette tip can be discharged. 20. The procedure is now complete.

[0220] In some embodiments, the IVF / ICSI platform may include labeling of patient samples. For example, to prevent sample confusion, the system may incorporate a labeling system to track multiple semen samples within a confined space and enable safe handling.

[0221] In some embodiments, the IVF / ICSI platform may include a liquefaction test. For example, the platform may use computer vision (e.g., microscopy) alone or in combination with pipetting and preparation protocols, as described herein, to assess whether semen agglutinations, clusters, or lumps have been broken down and whether the sample is liquid.

[0222] In some embodiments, the IVF / ICSI platform may include viscosity testing. For example, the platform may evaluate sample viscosity through periodic "thread" formation testing using pipetting operations, where thread length and breakage are assessed (e.g., macroscopic inspection) using optical and machine vision processes described herein, and the platform may be measured by video recording.

[0223] In some embodiments, the IVF / ICSI platform may include viscosity reduction techniques. For example, the platform may incorporate methods such as vigorous pipetting (mechanical), ultrasound, enzymes, or other means to reduce viscosity as needed.

[0224] In some embodiments, the IVF / ICSI platform may include processes that can perform overlay or underlay. For example, the platform may include an underlay that involves layering semen beneath a processing medium to facilitate swim-up.

[0225] In some embodiments, the semen preparation module can apply ultrasound to the sperm sample to facilitate the breakdown of protein structures within the seminal plasma. For example, an ultrasonic transducer can be used to deliver ultrasound to manipulate the protein structures and improve the viscosity and density of the sample, thereby improving overall sample preparation and pipetting performance, such as more robust and properly densely packed pipette tips. As shown in Figures 22 and 23, the pipette tip may have a custom shape according to the embodiments described herein to facilitate the sperm swim-up protocol. In some embodiments, the pipetting operation of the sample may include automatically optimizing the adjustment of the tilting mechanism, including pipette movement (e.g., upward and downward movement of the pipette, plane of movement, etc.), as well as tilting motion, degree of tilt, speed of tilt, duration of tilt, and other gradient dynamics, using the AI / ML process described herein. Additional processes, such as the application of protein-enriched medium to the sperm sample, may be further adjusted and optimized during the pipetting operation process using the AI / ML process described herein.

[0226] In some embodiments, multipurpose processing medium (MHM) or other culture or processing medium can be used by the semen preparation module as part of the preparation of sperm samples for pipetting.

[0227] In some embodiments, the IVF / ICSI platform includes a pipette-based double-layer or multi-layer system for semen preparation, which can reduce the number of preparation processes (and consumables) and avoid cross-contamination of seminal plasma. The process is initiated by drawing the processing medium into the pipette tip, and then, without releasing pressure, the semen can be continued to be drawn so that the semen is positioned at the bottom of the pipette tip with the processing medium above it. Residual pressure at the pipette tip prevents the double layer from leaking out of the open end of the pipette tip. The pipette tip can be held in place for a specified period to allow sperm to swim up into the medium layer. After the swim-up is complete, the semen layer, along with a specified volume of processing medium, can be discharged through the opening at the end of the pipette tip. The remaining processing medium layer, containing motile sperm, can be retained and transferred to a clean, sterile tube. The preparation can be used for both IVF and ICSI.

[0228] In some embodiments, the IVF / ICSI platform may include efficient separation of sperm from seminal plasma / body fluids. For example, the platform may be able to collect the final sample without any contamination from seminal plasma due to the precision of the robotic pipetting system.

[0229] In some embodiments, the IVF / ICSI platform may include optical evaluation of the swim-up layer. For example, the platform may use a photodetector or camera to monitor the swim-up process. Using the ML / AI described herein, the number of motile sperm in the swim-up fraction can be estimated, and the time required to collect a suitable sperm suspension can be determined. With the incorporation of ML / AI, the platform may "stop" the procedure, alert the operator, or continue with the next process of the procedure, which is sample collection.

[0230] In some embodiments, the IVF / ICSI platform may include a "combined pipette tip" (e.g., a 10 μL tip attached to the tip of a 1000 μL tip) that eliminates the need for multiple pipette sizes. This combined pipette tip can be used with other modules of the IVF / ICSI platform described herein.

[0231] In some embodiments, the IVF / ICSI platform can be scalable. For example, the platform can be expanded to accommodate the simultaneous processing of multiple semen samples based on clinical needs.

[0232] In one exemplary embodiment, computer vision, machine vision, or the like can be used to image the sample in the pipette and determine the validity of the volume, density, or any other characteristic of the pipetted sperm sample. The image of the pipetted sample can be processed using artificial intelligence, machine learning, or the like, as described herein, to test the sample against criteria for determining the quality of the sample. Continuing the example, if the pipetted sample is determined to be unsuitable by the automated AI / ML system, the pipetting process can be repeated, followed by further imaging and processing by the AI / ML process. If a satisfactory sample is not obtained after a specified number of pipetting attempts, the IVF / ICSI platform can issue a failure warning. The semen preparation module may also include an emergency stop activator (e.g., a button) within the user interface associated with the IVF / ICSI platform (2700) (see Figure 27). Warnings regarding unsatisfactory pipetting events can be issued within the IVF / ICSI platform, for example, on the IVF / ICSI platform user interface used by operators exchanging information with the IVF / ICSI platform. Alternatively, the warning may be issued to systems or platforms outside the IVF / ICSI platform, including, but not limited to, telecommunications platforms (e.g., sending SMS text messages, phone calls, videos of pipetting operations, machine-generated voice narration summarizing the pipetting failure, or any other message type), email platforms, electronic medical records, laboratory platforms, medical payers and / or reimbursement platforms, physician consultation platforms, or any other external platform type. Upon detecting a pipetting failure, the semen preparation module may generate a list of instructions that a human and / or robotic intervener can use to continue and / or resume the pipetting operation. Alternatively, the continuation or resumption of the pipetting operation may be triggered automatically according to a set of criteria stored by the IVF / ICSI platform.In some embodiments, the stored criteria can be further associated with selection rules relating to data concerning pipetting events. The unique data of a given pipetting event can be processed by the IVF / ICSI platform according to rules and an appropriate set of instructions selected to determine the next process and actions to be taken, based at least in part on what happened during the given pipetting event.

[0233] In some embodiments, the semen preparation module may include a biological residue waste container to which used or discarded pipette tips can be automatically attached. For example, within the semen preparation module, the preparation of an unprocessed semen sample can be initiated using a first pipette, such as by performing a washing routine. This same pipette can then be used to fill the container with a pre-prepared sperm sample, which can then be automatically discarded by the semen preparation module. The semen preparation module can then acquire a second pipette to handle the MHM and place it on top of the semen pre-placed in the container. Once the MHM is placed, the second pipette can be discarded, and a third pipette can be automatically acquired by the semen preparation module. The third pipette can then be used to selectively acquire material from the top of a layered sample, which can then be automatically discarded after use.

[0234] In some embodiments, the semen preparation module may include labeling of samples, instruments, and devices to enable automated imaging and tracking of each process performed by the semen preparation module. For example, a camera within the semen preparation module can image, record, store, and analyze samples, containers, pipettes, instruments, etc., including the clinical sequence, operation, performance, and results of each process, such as the degree of separation of liquids, e.g., in a bilayer. Labeling can also be used to track the person from whom the semen sample originates, such as a physician, laboratory professional, office, company, insurer, payer, or any other party further related to sperm sample processing. Data derived from the completion of each process performed in the semen preparation module can be recorded and stored within the IVF platform or within a platform associated with the IVF platform, such as electronic medical records or other data repositories.

[0235] In some embodiments of the present invention, the IVF platform described herein may include, for example, a system for the fully autonomous preparation of semen for ART or IUI.

[0236] In some embodiments of the present invention, the IVF platform described herein may include stratified automated semen preparation that mimics the human interphase process but requires a high degree of efficiency in separating seminal plasma.

[0237] In some embodiments of the present invention, the IVF platform described herein has the ability to automatically process semen samples from multiple sources within a confined space secured by a warning system, thereby avoiding confusion errors.

[0238] In some embodiments of the present invention, the IVF platform described herein may include an automated built-in system for reducing viscosity and facilitating liquefaction.

[0239] In some embodiments of the present invention, the IVF platform described herein may include a centrifuge-free robotic system that enables fully autonomous preparation of sperm for ART or IUI from all male factor pathologies without swim-up. In some embodiments, such robotic processing by the IVF platform can replace the male pathologist or embryologist who prepares semen samples.

[0240] In some embodiments, the IVF / ICSI platform may include a robotic sperm preparation system for human clinical and / or veterinary purposes that does not use sperm swim-up, centrifugation, or microfluidics. This robotic sperm preparation system can be used for the automated preparation, selection, and processing of single sperm cells from any mammal. The robotic sperm preparation system includes single sperm collection from a seminal plasma / culture suspension, or directly from a culture droplet without seminal plasma, after horizontal swimming from a droplet to culture medium. The robotic sperm preparation system can be used for any mammal having functional sperm in the testes or seminal fluid, including, but not limited to, azoospermic men from whom sperm can be surgically retrieved from the epididymis or testes, including those with the lowest sperm counts.

[0241] In some embodiments, a robotic sperm preparation system may include robotic cell isolation, droplet pipetting, use of a mineral oil-submerged culture medium pathway in a dish or on a slide, and integration of single motile sperm selection with sperm immobilization and injection techniques. The robotic sperm preparation system process may be organized by a series of autonomous commands or protocols as described herein, and can be combined to ensure efficient and precise execution and achieve complete autonomy. Parallel sperm preparation from different males can be achieved by combining multiple systems (e.g., semen preparation modules as described herein) horizontally, stacked, and / or mobile, such as robot-controlled semen preparation modules that can move relative to each other and to other modules of the IVF / ICSI platform as needed to facilitate parallel processes on the IVF / ICSI platform. The yields and protocols of the IVF / ICSI platform can be adjusted to allow IUI, IVF, or ICSI as desired. The robotic sperm preparation system may not require the use of microfluidics, but microfluidic disposable supplies can be incorporated as needed. Robotic sperm preparation systems may not require centrifugation or swim-up stacking systems.

[0242] In some embodiments, the robotic sperm preparation system may include the following: Connect to a semen analysis system or perform semen analysis directly. It utilizes pipetting operations from multiple robotic pipettes positioned at arbitrary angles. Microscopic observation of sperm at different locations within the system, and / or direct observation with a digital camera, are performed. Multiple observations of the preparations can be performed simultaneously. This device is used to pipette semen and surgical fluids obtained from sperm aspiration from the epididymis or testis. A temperature-controlled receiving section for surgical tubes can be integrated. Prepare pipette droplets using a dish and, for example, 100 nL to 100 microliter droplets. For coating the microdroplets, a larger volume of oil coating in the range of 1 to 5 mL can be used. The IVF / ICSI platform fabricates round droplets and arbitrary linear and nonlinear bodies according to any programmable design commanded by the platform, and connects the droplets. For example, a system that optically and continuously measures sperm migration can be used to block horizontal sperm migration. A customizable culture medium dish is prepared using a programmable system commanded by the IVF / ICSI platform.

[0243] In some embodiments, the robotic sperm preparation system may include sperm processing: (a) non-surgical sperm collection. The semen sample can be evaluated using semen analysis after liquefaction. If liquefaction is not present, it can be initiated using vigorous robotic pipetting or by mixing with a proteolytic enzyme such as chymotrypsin. In the latter case, depending on the pipetting protocol, the enzyme can be rapidly removed from the enzyme solution into small culture medium droplets by a single robotic sperm pipetting operation until the enzyme is no longer present. The naturally liquefied sample can be mixed with buffered culture medium, and the robotic arm can collect motile sperm one by one after motility and morphological evaluation using AI / ML and optics, for example, as described herein. AI / ML and optics can be applied when liquefaction is not present and enzymes are being used. The system can collect enough sperm to provide secondary sperm identification and selection (SiD) analysis after the sperm have been mixed with droplets containing, for example, a 5-10% solution of PVP under oil in a prepared ICSI dish. This can, for example, be included as a final step before robotic sperm selection and immobilization using robotic ICSI.

[0244] In some embodiments, the robotic sperm preparation system may include sperm washing and selection. The system may optionally use SiD technology, supplemented with morphological evaluation and segmentation analysis, to rapidly select optimal sperm during initial preparation, after which the sperm are injected into droplets containing PVP to reduce their immobilization and advancement for ICSI. Seminal plasma can be removed by repeatedly robotically collecting single sperm in groups of 5–300, for example, into a fine pipette, and continuously pipetting the sperm through buffer medium. Multiple pipettes can be used simultaneously. For surgical preparation, larger clumps of cells and tissues can be removed using larger pipettes. The system can operate with one robotic pipette or several robotic pipettes simultaneously. Robotic systems using sperm collection without horizontal sperm preparation can be considerably faster than standard manual procedures. The horizontal sperm preparation procedure can also be equipped with SiD AI to time the horizontal swim-out to the optimal concentration in a culture medium without seminal plasma. The timing may vary from sample to sample, depending on sperm advancement and velocity. The optimization of culture medium droplet size and placement, as well as the linking of culture medium pathways, may depend on the quality of the semen sample. The configuration shape may vary and can be optimized through experience. Each droplet and pathway configuration can be optimized using the optics of the AI / ML and IVF / ICSI platforms.

[0245] In some embodiments, the robotic sperm preparation system is fully automated and can utilize advanced robotic mechanisms and 3D motorized systems, including integration with cameras and computer vision to facilitate autonomous control of the system, to ensure precise movement throughout the entire process.

[0246] In some embodiments, the robotic sperm preparation system can autonomously manage autofocus, zoom in, and zoom out using the software and hardware systems of the IVF / ICSI platform for clear and accurate imaging at all stages.

[0247] In some embodiments, sperm tracking within a robotic sperm preparation system can be a continuous process, allowing for real-time monitoring of sperm motility and location throughout the procedure. Optimal sperm preparation concentration can be maintained by blocking horizontally positioned pathways using a blunt pipette tip that automatically blocks the sperm pathway.

[0248] In some embodiments, a robotic sperm preparation system can identify and track various components of sperm, including the tail, acrosome, head, and midpiece. This segmentation process ensures the precise processing of live, motile sperm.

[0249] In some embodiments, the robotic sperm preparation system may include sperm re-evaluation. For example, after selection, the semen preparation module may provide options for further SiD evaluation either before or after immobilization to confirm the suitability of the selected sperm.

[0250] In some embodiments, the robotic sperm preparation system may include data management and record keeping, for example, by incorporating automated patient identification and data recording capabilities, and ensuring comprehensive documentation and quality control of each procedure.

[0251] In some embodiments, the robotic sperm preparation system can be housed in a space where temperature and humidity are controlled by HEPA and carbon filtration.

[0252] In some embodiments, the robotic sperm preparation system may include, but not limited to, quality control of operational functions (disposable supplies, bodily fluids, pipette tips, discarded disposable bags, etc.) controlled by a fully automated quality control process.

[0253] In some embodiments, the robotic sperm preparation system may include command-based operating systems and methods. For example, a single command may initiate a series of predetermined events, streamlining the process and reducing the possibility of human error. This feature improves efficiency and reproducibility and can also be interrupted by a technician or a professional embryologist or male pathology technician. Interphase can be adapted to digitally control aspects such as channeling, sperm concentration determination, focusing, sperm positioning, lens or camera changes, stage and / or camera or lens movement, microtool movement, light adjustment, insertion of different microscope systems, droplet location determination, or any other operation of the IVF / ICSI platform.

[0254] Figure 61 shows an exemplary robotic sperm preparation 6100 that can be performed by the IVF / ICSI platform by pipetting single sperm into groups using a semen / culture mixture and sperm tracking system and diluting seminal plasma using a droplet washing system. In some embodiments, the semen (lower layer) can be mixed with the culture medium (upper layer) using a robotic pipette by using a collection bottle 6102, for example, by collecting the ejaculate into a dry bottle, or by collecting the ejaculate into a bottle filled with culture medium (e.g., 2-5 mL). The robotic mechanism of the IVF / ICSI platform can mix the semen (sperm and seminal plasma) with the culture medium. Droplets of culture medium / seminal plasma / sperm can be pipetted to the bottom of a dish 6104 by a robotic arm. Subsequently, the droplets are covered with mineral oil by the robot. The semen / culture mixture can be placed on a microscope stage 6106, and sperm can be visualized using optical and imaging systems as described herein. For example, sperm can be collected using multiple robotic pipetting arms equipped with thin glass or plastic pipette tips (6110).

[0255] Referring to Figure 62, collection tubes 6200 can be filled with biopsy material from the testes or epididymal aspirates (shadowed areas) (6202, 6204, 6206). The robotic mechanism of the IVF / ICSI platform can place the tubes on temperature-controlled shelves. Collection vials containing ejaculate 6204, 6206 (shadowed) from individual patients or donors can be placed in individually controlled and monitored compartments within the modular system of the IVF / ICSI platform. Each compartment processing specimens from a single patient can perform the robotic work required to complete sperm preparation, or perform horizontal swim-out under mineral oil using a microdroplet and micropathway medium system. In some embodiments, the compartment system can be modular and expandable. Each compartment can be environmentally isolated with its own material supply and disposable area and may include a label recognition system.

[0256] Referring to Figure 63, a robotic sperm preparation method is shown that uses horizontal swim-out, without swim-up, centrifugation, or microfluidics (6300). The robotic system of the IVF / ICSI platform can prepare the dish and monitor the swim-out. Once a predetermined number of motile sperm are obtained, the pathway can be blocked by creating oil crosslinks within the pathway using the robot. For example, 1-10 microliter droplets of seminal plasma as single or multiple droplets using a circular or star shape can form connections to medium droplets without seminal plasma using a medium pathway system that allows sperm to move through the medium pathway under the oil. Sperm can be collected individually or in groups containing multiple sperm. Other droplets containing PVP can be connected to the droplet pathway system to slow down sperm that have swum from the seminal plasma droplets. Sperm that have reached the PVP can be selected, fixed, and prepared for ICSI. Droplets containing oocytes can be present in the same dish. Medium droplets and medium-filled pathways 6302 can be pipetted before being covered with an oil film. Semen droplets 6304 can be arranged in single or multiple configurations to improve yield when sperm quality is poor. The robotic mechanism of the IVF / ICSI platform can be programmed to pipette multiple culture medium type configurations. A heating element can be placed under the seminal plasma droplets or pathway to enhance swim-out. In an exemplary configuration, droplets and pathways can be connected to PVP droplets 6306. In another exemplary configuration, culture medium droplets can be added for the placement of ICSI-ready eggs 6308. Testicular aspirates and other samples with poor counting and motility can be pipetteted with larger, thinner droplets connected to multiple culture medium pathways to increase yield. This can be combined with the robotic mechanism used to physically obtain sperm.

[0257] In some embodiments, the semen preparation module can move sperm from liquefied semen or semen mixed with some culture medium by visualizing the sperm in an automated manner using a pipetting system and robot, or multiple robots, which use a thin capillary or capillary-like tip to quickly focus on the sperm and remove the sperm by aspiration. In some embodiments, these robotic processes can be combined with DL / CV-based sperm selection software as needed. The sperm extracted from the semen sample can be deposited into culture medium droplets with slightly increased viscosity using dextran, PVP, or an equivalent viscous substance. This facilitates pipetting operations to the final process. After such processes, the sperm can be deposited, for example, in culture medium or in a PVP solution ready for ICSI for later use.

[0258] In some embodiments of the present invention, the IVF / ICSI platform described herein may include automated semen preparation without stratification or centrifugation.

[0259] In some embodiments of the present invention, the IVF / ICSI platform described herein may include such systems that are potentially applicable to any semen or surgically recovered sperm sample.

[0260] In some embodiments of the present invention, the IVF / ICSI platform described herein may include an automated built-in system for reducing viscosity and facilitating liquefaction.

[0261] In some embodiments of this disclosure, the semen preparation module may include a robotic system for semen preparation (swim-up), collection, and transfer to an ICSI dish. In some embodiments of this disclosure, the semen preparation module may include AI / ML for automated detection, identification, and classification. In some embodiments of this disclosure, the semen preparation module may include AI / ML for automated measurement and testing. In some embodiments of this disclosure, the semen preparation module may include AI / ML for optimization. In some embodiments of this disclosure, the semen preparation module may include AI / ML for prediction. In some embodiments of this disclosure, the semen preparation module may include AI / ML for selection / grading. In some embodiments of this disclosure, the semen preparation module may include AI / ML for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the semen preparation module may include AI / ML for system configuration and control. In some embodiments of this disclosure, the semen preparation module may include fully autonomous AI / ML. In some embodiments of this disclosure, the semen preparation module may include optical and machine vision components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include robotic processing systems and processes. In some embodiments of this disclosure, the semen preparation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include sperm preparation components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the semen preparation module may include laser components, systems, and processes.

[0262] In some embodiments of the present invention, the IVF / ICSI platform described herein may include an oocyte preparation module that is fully automated and uses robotic mechanisms for processing and moving materials, including biological specimens, and is connected to the network infrastructure described herein for remote control of the operation of the oocyte preparation module as one component of a fully automated robotic IVF / ICSI platform.

[0263] In some embodiments, follicle-stimulating hormone (FSH) can be administered to a patient in a medical setting to stimulate the ovaries and promote the growth of multiple follicles, and an ultrasound-guided transvaginal procedure with aspiration of the growing follicles can be used while the patient is under sedation. The follicles are punctured using a fine needle and the fluid inside is aspirated into the tubule. A mature antral follicle at ovulation may be about 25 mm in diameter and contain about 50 million granulosa cells and about 7 ml of follicular fluid. Follicular fluid is often opaque yellow and contains, among other factors, hormones (e.g., estrogen, progesterone, androgens), growth factors, cytokines, metabolites (e.g., glucose, pyruvate, lactate), ions (e.g., sodium, potassium, calcium), and proteins (e.g., albumin, transferrin). The cellular components of follicular fluid may include granulosa cells. As the (antral) follicle develops, different classes of functionally distinct granulosa cells are generated depending on the position of the granulosa cells relative to the oocyte. The cells closest to the oocyte are cumulus granulosa cells, while mural granulosa cells are further away from the oocyte and line the follicular wall. In response to ovulation induction, cumulus cells secrete an extracellular matrix (and therefore hyaluronidase-sensitive) composed mainly of hyaluronan, which causes the cumulus cells to swell in a process called myelogenesis. The myelinated or swollen cumulus mass is translucent and appears different from mural granulosa cells, which maintain their sheet-like, tightly woven form with a darker appearance and are therefore visually distinguishable from the cumulus even with the naked eye.

[0264] In some embodiments, the biological material obtained from a patient during oocyte retrieval may include the oocyte-cumulus-coronal radiata complex (OCCC) or cumulus-oocyte complex (COC). The coronal radiata cells are a layer of cells in direct contact with the zona pellucida, a cell-free glycoprotein portion or "shell" surrounding the oocyte via cytoplasmic processes. During cumulus expansion, the coronal radiata cells can be separated from the zona pellucida, but they are still recognizable as a distinct layer of cells surrounding the oocyte. They do not undergo mucinization and usually appear darker than the cumulus cells, facilitating the visual identification of the oocyte. Overstimulated ovaries have increased vascular distribution, and follicles are more prone to bleeding when punctured by the retrieval needle. This can result in a mixture of blood and follicular fluid.

[0265] In some embodiments, oocytes isolated from follicular fluid may be at different stages of nuclear maturation, e.g., fully mature (metaphase II or MII of meiosis with a first polar body), intermediately mature (metaphase I of meiosis, without a polar body), and immature (prophase I of meiosis containing a large nucleus called a nucleus vesicle or GV). Depending on the level of nuclear maturation, the cumulus mass may appear in different forms. It is swollen and translucent in MII oocytes, somewhat darker and more densely organized in MI oocytes, and extremely dark, dense, and small in GV oocytes. Nuclear maturation can be estimated based on the appearance of the cumulus and corona radiata cells.

[0266] In some embodiments, the IVF / ICSI platform can autonomously robotically perform oocyte discovery and isolation, replacing conventional oocyte retrieval procedures performed by a human operator. Conventional oocyte retrieval procedures performed by a human operator generally follow the processes listed in Table 2 below.

[0267] [Table 2]

[0268] In some embodiments, embryologists can examine the follicular fluid aspirate under a stereomicroscope to identify and isolate COCs. The cumulus oophorus of the oocyte can be incised using a subcutaneous needle to remove blood clots or diseased-looking cells, or its size can simply be reduced before incubation.

[0269] In some embodiments, pipetting in the context of oocyte retrieval may involve aspirating, discharging, and transferring individual or multiple COCs from container to container. Pipetting is fundamental to IVF laboratory techniques and can be performed in a sterile manner without generating bubbles that can be destructive, result in cell loss, or generate potentially infectious aerosols. Different types of pipettes, including but not limited to Pasteur pipettes, Eppendorf pipettes, capillary tips, or any other type of pipette, can be used in oocyte retrieval procedures by the automated robotic pipetting systems and methods of the IVF / ICSI platforms described herein. Pasteur pipettes are made of borosilicate glass and typically come in two lengths (short, i.e., 5.75 inches, and long, i.e., 9 inches) and are used with rubber pipette valves to transfer smaller amounts of body fluid, with or without cells. These pipettes can be "pulled" over a flame to create a very narrow opening for handling oocytes and embryos. An Eppendorf pipette is a device that uses an Eppendorf tip to aspirate and dispense precise (usually low) volumes, featuring a piston and spring-loaded conical tip, a single pathway, and adjustable volume (in increments of 1 to 1000 μL within a specific range). A capillary tip refers to the pipette tip used in conjunction with a capillary pipette. The tip is made from flexible, medical-grade plastic to prevent scratching of plastic petri dishes. Tips are manufactured with different inner diameters ranging from 75 μm to 600 μm, with the most commonly used sizes being 155–200 μm for oocyte detachment and processing of oocytes and embryos, and 300 μm for blastocyst processing.

[0270] The automated IVF / ICSI platform can utilize robotic mechanisms for the isolation, processing, and transfer of oocytes with COCs from the follicular fluid. The supernatant of the follicular aspirate can be poured into a dish, and the follicular aspirate can be placed on an electric stage of an inverted microscope. The dish can be automatically scanned in a predetermined pattern using computer vision, in combination with AI / ML and / or computer vision and optics as described herein, to identify the COCs. Once identified, the COCs can be automatically collected with a pipette, washed, and transferred to a new dish containing fresh processing medium.

[0271] In some embodiments, exemplary simplified sequences for oocyte isolation are shown below, each of which can be autonomously and robotically executed by the IVF / ICSI platform described herein.

[0272] The processing medium and a "washing" dish (e.g., 1 x 35 mm) with an oil film (to prevent evaporation) can be received by the IVF / ICSI platform and placed on the stage of an inverted microscope or some other type of microscope equipped with a "dish holder" (e.g., a rectangular component with notches to fit one 60 mm and 2 x 35 mm culture dish).

[0273] The IVF / ICSI platform can use computer vision to initiate a sequence involving stage movement, for example, by scanning a dish in a predetermined zigzag pattern from top to bottom.

[0274] AI / ML can identify the cumulus mass containing oocytes. Each aspirate can contain zero to multiple oocytes. This depends in part on the aspirate method used by the surgeon and the size of the follicles being aspirated. Individual follicles can be aspirated into each tube, or multiple follicles can be aspirated into one tube simultaneously. The oocyte-to-tube ratio can be other than 1:1, and there may be tubes with more aspirate fluid and washing medium than oocytes, or multiple oocytes may be in one tube. The robotic mechanism of the IVF / ICSI platform described herein can identify different contents and identify oocytes. In some embodiments, processes 3 and 4 may or may not be performed in parallel. For example, while the dish is being scanned, computer vision, optics and / or AI / ML can be used to search for and identify COCs, or the process can stop scanning at a given location and interact with computer vision, optics and / or AI / ML, and / or the IVF / ICSI platform can acquire images of the follicular fluid during the scanning process and, after the scanning is complete, query computer vision, optics and / or AI / ML to determine whether COCs are present.

[0275] Once identified, the pipette held in the microtool holder can be lowered into the follicular fluid dish and positioned directly adjacent to the cumulus mass. In one embodiment, the pipette tip can be lowered at a certain distance from the COC, and once inside the liquid, it can approach the COC and then begin aspirating.

[0276] Next, negative pressure is applied, allowing the cumulus mass to be aspirated into a pipette along with some body fluid.

[0277] The volume of body fluid can be precisely controlled based on the physical limits outside the mass. Once the entire mass enters the pipette, aspiration can be stopped. The pipette can then be lifted from the dish and left stationary.

[0278] The stage can then be moved toward a first dish containing the processing medium.

[0279] When it reaches the designated location, the pipette can be lowered into the washing dish, positive pressure can be applied, and the cumulus mass / follicular fluid can be discharged into the culture medium. The volume can be precisely controlled so that the positive pressure stops when the entire mass has exited the pipette.

[0280] Next, the pipette can be lifted from the dish and kept still.

[0281] In some embodiments, the oocyte preparation module can receive a follicular fluid sample and automatically position the sample for examination with a microscope, computer or machine vision, or any other imaging device, in order to identify, locate, analyze, and evaluate the cumulus-oocyte complex (COC). In one exemplary embodiment, the sample can be examined within the oocyte preparation module using an inverted microscope, a digital microscope, or any other type of microscope. In some embodiments, the inverted microscope may include components and adapted robotic mechanisms used as part of an automated ICSI procedure, as described herein, and may be equipped with a digital microscope both below and above the plate, dish, or other type of container containing the sample. The follicular fluid sample can be automatically scanned using computer vision, machine vision, etc., in combination with AI / ML to perform COC identification. This imaging in combination with AI / ML can enable the distinction of cell types, such as those of hemocytocytes derived from the COC.

[0282] In some embodiments, during a specific COC (Clinical Occlusion) step, the oocyte preparation module can autonomously robotically position the dish, receptor, or container containing the follicular fluid sample onto a motorized stage, plate, or other surface to move to the stage or plate where the follicular fluid sample is located. Figures 36 and 37 show exemplary forms of dish holders that can be used by the oocyte preparation module to hold the sample dish, receptor, or container (3600, 3700). When imaging and AI / ML procedures are performed on the follicular fluid sample, the oocyte preparation module can automatically adjust the position of the dish, receptor, or container in which the follicular fluid sample is placed to optimize, for example, the angle, height, or portion of the sample when the sample is being imaged. The stage, plate, or other surface can move along any axis of movement and can rotate along any axis or plane of operation.

[0283] In some embodiments, imaging of the sample within the oocyte preparation module can be performed independently of or through a microscope. For example, a set of imaging equipment can be used to evaluate and analyze the arrangement of certain instruments, such as the position of dishes, receptors, or containers used within the oocyte preparation module, independently of a microscope. In some embodiments, imaging can also be performed through the optical section of a microscope, for example, by fitting a camera or multiple cameras to a microscope observation device such as binoculars. In some embodiments, during imaging, the oocyte preparation module can automatically adjust the frequency, intensity, angle, distance, or any other factor of the artificial illumination used to image the sample. The adjustment of the frequency, intensity, angle, distance, or any other factor of the artificial illumination used to image the sample can be at least in part based on the AI / ML process used to evaluate the images obtained from the follicular fluid sample by the oocyte preparation module.

[0284] In some embodiments, for example, by using an inverted microscope in combination with imaging and AI / ML as described herein, once at least one COC is identified, a pipette can be lowered vertically into the follicular fluid sample to extract at least one COC and transfer the selected COC to a second dish, receptor, or container in an oocyte preparation module containing culture medium. The dish, receptor, or container in which the COC is placed may be on or near a temperature-controllable plate, such as providing the dish, receptor, or container with a constant 37 degrees Celsius (or some other target temperature) environment during the COC washing and preparation process. If more than one COC is identified in the sample during this process, the oocyte preparation module may further separate the COCs into additional dishes, receptors, or containers to provide a single COC per segmented sample, or some other target number of COCs per segmented sample.

[0285] In some embodiments, during the COC identification and separation stage, the oocyte preparation module can automatically apply compounds to the sample to facilitate COC separation and extraction. For example, if the imaging and AI / ML procedures of the oocyte preparation module detect the presence or possibility of blood or blood coagulation in the follicular fluid sample, the oocyte preparation module can robotically select the amount of heparin or other anticoagulant and add heparin to the follicular fluid sample to facilitate COC extraction. The type of anticoagulant, the amount of anticoagulant, the timing of the addition of the anticoagulant to the follicular fluid sample, and other factors can be determined at least partially automatically using the imaging and AI / ML processes of the IVF / ICSI platform described herein.

[0286] In some embodiments, once the COC is autonomously and robotically placed in a dish, receptor, or container containing culture medium, the oocyte preparation module can move the COC to a new dish, receptor, or container where it performs a series of autonomous and robotic washes of the COC. Following the washing of the COC, the dish, receptor, or other container containing the COC can be automatically transferred to an incubator.

[0287] In some embodiments, after the culture period of COC, the oocyte preparation module can autonomously and robotically initiate the oocyte detachment process.

[0288] In some embodiments, the automation used herein includes robotic mechanisms and AI / ML-assisted processes so that processes in oocyte detachment procedures, which normally require human operators, can be performed by an intelligent robotic system of the IVF / ICSI platform. Table 3 outlines such processes involved in conventional oocyte detachment.

[0289] [Table 3]

[0290] As the (antral) follicle develops, different classes of functionally distinct granulosa cells are generated depending on the position of the granulosa cells relative to the oocyte. The cells closest to the oocyte are called cumulus cells (or cumulus follicles). Immediately surrounding the oocyte are corona radiata cells, which are in direct contact with the zona pellucida via cytoplasmic processes. During cumulus expansion (after ovulation induction), the corona radiata cells separate from the zona pellucida (the processes are mostly retracted), but they are still recognizable as a distinct layer of cells surrounding the oocyte. They do not undergo myxogenesis and usually appear darker than the cumulus cells, facilitating the visual identification of the oocyte.

[0291] In some embodiments, oocytes isolated from follicular fluid may be at different stages of nuclear maturation, namely, fully mature (metaphase II or MII of meiosis with a first polar body), intermediately mature (metaphase I of meiosis, without a polar body), and immature (prophase I of meiosis containing a large nucleus called a nucleus vesicle or GV). Depending on the level of nuclear maturation, the cumulus mass may appear in different forms. It is swollen and translucent in MII oocytes, somewhat darker and more densely organized in MI oocytes, and extremely dark, dense, and small in GV oocytes. Nuclear maturation can be reasonably estimated based on the appearance of the cumulus and corona radiata cells, but the developmental potential of the oocyte cannot be assessed by this method.

[0292] In some embodiments, the cumulus-coronal-radiata complex can be removed to allow visualization and micromanipulation of the oocyte before sperm injection into the oocyte. Complete removal of the cumulus and coronal-radiata cells is called detachment. This is achieved enzymatically and mechanically. The cumulus cells can be dissociated using the enzyme hyaluronidase. This is possible because the swollen cumulus is an extracellular matrix rich in hyaluronan (HA). On the other hand, the coronal-radiata cells do not undergo mylulation and can therefore be removed mechanically. Mechanical removal can be achieved by repeated aspiration / discharge using a pipette with a small bore diameter, e.g., a 155-200 μm capillary tip or a Pasteur pipette stretched by hand.

[0293] In some embodiments, pipetting in the context of detachment may involve the aspiration, efflux, and transfer from container to container of individual or multiple OCCCs and oocytes. Pipetting is fundamental to all IVF laboratory techniques and can be performed in a sterile manner without generating bubbles that can be destructive, result in cell loss, or generate potentially infectious aerosols. Different types of pipettes can be used during the detachment procedure, including, but not limited to, the following: Pasteur pipettes: Made of borosilicate glass, for example, come in two lengths (short, i.e., 5.75 inches, and long, i.e., 9 inches), and are used with rubber pipette valves to transfer smaller amounts of body fluid, with or without cells. These pipettes can be "pulled" over a flame to create extremely narrow holes for processing oocytes and embryos. Eppendorf pipette: A device equipped with a piston and spring-loaded toothed cone, a single pathway and adjustable volume (in units of 1 to 1000 μL with a specific range), used in conjunction with an Eppendorf tip to aspirate and dispense precise (usually low) volumes. Capillary tip: A pipette tip used in conjunction with a "stripper" or "EZ-grip" pipette. The tip is made from flexible medical-grade plastic to prevent scratching of plastic petri dishes. The tip can be manufactured with different inner diameters ranging from 75 μm to 600 μm, with commonly used sizes being 155-200 μm for oocyte detachment and processing of oocytes and embryos, and 300 μm for blastocyst processing.

[0294] In some embodiments, exemplary hardware configurations for peel-related processes may include, but are not limited to, the following instruments that can be integrated within the IVF / ICSI platform described herein. Inverted microscope (Olympus® IX81) Stage movement control device (Prior H117P2IX) Microscope Dino-Lite Edge (5mp series) ArduCam (IMX477 12MP) Stage heating & control device (TokaiHit) Fine-tuning device (Eppendorf; TransferMan 4r) A range of 12,500 μm for each axis. Maximum speed 10,000μm / sec Microinjector (Narishige IM-21) 10 μL per rotation Total 400 μL W127-147×D56×H78mm Microtool holder (HI-7) W140×D8×H78mm 2-Stepper Motor (5PCS Nema 17) LCPlanFI 20x Olympus UPlanFLN 4x Olympus Motor control device BTT SKR Mini E3 v3.0

[0295] In some embodiments, the automated oocyte, AI / ML-assisted identification and isolation system of the IVF / ICSI platform can minimize and / or eliminate the need for a skilled embryologist for oocyte detachment in preparation for ICSI or cryopreservation.

[0296] In some embodiments, during the detachment process, the specimen held in a dish can be imaged using a stereomicroscope, inverted microscope, or some other type of microscope, including a dish with multiple depressions, e.g., a dish having four or more depressions within it. The first depression may contain an enzyme that allows for the removal of cumulus cells from the oocyte. The oocyte may be drawn in from the depression into the pipette and then discharged from the pipette using a pipette, and this process can be repeated, thereby removing the cumulus and corona radiata cells via enzymatic action and by the mechanical force of the pipette drawing in and discharging the oocyte in an automated manner using robotic processing. In one example, the oocyte preparation module may also accelerate detachment by automatically using a pipette or other device to agitate the fluid in which the COC is held. As the detachment process progresses, the degree of corona radiata cell removal can be periodically or continuously evaluated and analyzed using the imaging and AI / ML processes described herein, and it can be shown that the corona radiata cell removal process can be terminated when a target endpoint is reached or exceeded. Because prolonged exposure to enzymes can damage oocytes, the use of imaging and AI / ML processes can facilitate shorter exposure times and reduce stress on the oocytes by enabling rapid identification of completed preparations. In some embodiments, the oocyte preparation module can use a single-cell holder with a standard needle holder 3800 using an adapter 3800, as shown in the exemplary embodiment of Figure 38.

[0297] Figure 39 shows the ICSI dish adapter 3900, which can be coupled to the oocyte preparation module and used thereby (3900).

[0298] Figure 40 shows the mechanical barrier 4000 that interacts with the organic compound.

[0299] In some embodiments, once corona radiata cells have been properly removed from the COC, the oocyte preparation module can autonomously and robotically initiate the washing process using several indentations within a dish, receptor, or container for washing. Following washing, the oocyte preparation module can automatically assess the maturity of the oocyte using the imaging and AI / ML processes described herein. In one embodiment, the imaging and AI / ML processes can evaluate the morphology of the oocyte to assist in determining maturity, for example, whether the oocyte has polar bodies. If the imaging and AI / ML processes determine that polar bodies are present, the oocyte can be considered mature by the IVF / ICSI platform and deemed eligible to proceed to the fertilization module of the IVF / ICSI platform described herein. If the imaging and AI / ML processes determine that polar bodies are absent, the oocyte can be considered immature by the IVF / ICSI platform and automatically returned for further incubation, and once the further incubation cycle is complete, the imaging and AI / ML processes can perform a subsequent maturity assessment round. In some embodiments, continuous monitoring over time can be used for evaluation.

[0300] In some embodiments, the oocyte preparation module and incubation module can be operably coupled to components of an intelligent automated system of the IVF / ICSI platform. The oocyte preparation module may be responsible for oocyte retrieval, identification, classification, measurement, testing, optimization, prediction, selection / grading, and processing, performing these tasks using AI / ML processes, robotic processing systems, and advanced microscopy systems as described herein. In one example, once the oocyte preparation module has completed the task of identifying and retrieving oocytes, the oocytes can be automatically transferred by robot to the incubation module. The incubation module may include components and systems for incubation, sensor components for monitoring conditions within the module, and sample management components for processing oocytes during the incubation process. The incubation module may employ AI / ML processes for automated measurement and testing, optimization, prediction, selection / grading, and system configuration and control, and may also include advanced microscopy, imaging, and optical systems, including robotic processing systems and computer and machine vision systems. In some embodiments, the oocyte preparation module and the incubation module can work together to ensure that oocytes are properly prepared, processed, and incubated. The use of AI / ML processes in both modules automates these processes, reduces the possibility of human error, and increases the efficiency and effectiveness of the IVF procedure. The integration of these modules can enable a streamlined and efficient process from the initial preparation of oocytes to their incubation. This integration can be facilitated by the use of interconnected robotic IVF modules, which enable the automated transfer of materials and data between modules and ensure that oocytes are processed and prepared in a consistent and controlled manner necessary for the success of the IVF procedure.

[0301] In some embodiments, as part of imaging and AI / ML processes used to determine the morphology of the oocyte and, for example, the presence or absence of polar bodies, a three-dimensional reconstruction model can be constructed to represent the complete physical entity of the oocyte, in contrast to being limited to two-dimensional images obtained, for example, through microscopic images. In one example, a three-dimensional image of the oocyte can be constructed through images taken from multiple angles of the oocyte, such as images created by physically moving the oocyte and acquiring images from multiple sides of the oocyte. Observation of multiple sides of the oocyte can be achieved by automatically manipulating the oocyte with an oocyte preparation module to observe different sides of the oocyte, or by physically moving an imaging device, such as a microscope, around a stationary oocyte. In another example, a three-dimensional image of the oocyte can be based at least in part on oocyte inference data and predictive modeling. Such inference and predictive modeling can be based in part on prior data derived from oocyte imaging performed by an IVF / ICSI platform. In one embodiment, optical coherence tomography (OCT), optical coherence microscopy (OCM), near-infrared tomography, or any other technique may be used for oocyte imaging by the oocyte preparation module. OCT can be used to assess the maturity of oocytes by the IVF / ICSI platform. By providing a three-dimensional image of the oocyte, OCT can identify the presence and location of polar bodies, which are structures indicating oocyte maturity. As described herein, this information can be used to guide the ICSI process and ensure that the needle is introduced into an ideal position (as used herein, “fertilization process,” “fertilization system,” “fertilization,” etc., include ICSI, ITEM on ICSI, and all related ICSI systems, processes, and protocols).In some embodiments, the OCT system of the IVF / ICSI platform, used in conjunction with the AI / ML system of the IVF / ICSI platform, can be used to capture planar images of oocytes throughout their development, enabling more detailed tracking of their maturation, and to visualize multiple oocytes simultaneously, providing a more efficient method for assessing oocyte maturation than conventional methods performed by operators.

[0302] In some embodiments, the oocyte preparation module can use polarized and / or polarized OCT to automatically identify the presence or absence of meiotic spindles, define the best placement of oocytes during injection, and assess membrane integrity to enable the identification of successful injections. In some embodiments, oocyte maturity can be measured at least in part on the automatic identification of meiotic spindles using imaging and AI / ML processes described herein. In some embodiments, meiotic spindle testing can also be used to form predictive algorithms to help determine whether an oocyte is properly maturing and the probability of when it can mature in further incubation.

[0303] In some embodiments of this disclosure, the oocyte preparation module may include oocyte retrieval components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for automated detection, identification, and classification. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for automated measurement and testing. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for optimization. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for prediction. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for selection / grading. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the oocyte preparation module may include AI / ML for system configuration and control. In some embodiments of this disclosure, the oocyte preparation module may include fully autonomous AI / ML. In some embodiments of this disclosure, the oocyte preparation module may include optical and machine vision components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include robotic processing systems and processes. In some embodiments of this disclosure, the oocyte preparation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the oocyte preparation module may include enzymatic oocyte detachment processes, systems, and components. In some embodiments of this disclosure, the oocyte preparation module may include advanced microscopy systems and components.

[0304] This disclosure provides an automated ICSI platform that automatically performs ICSI using robotic mechanisms and AI / ML, including machine vision. The system aims to improve the consistency and success rate of ICSI procedures. The automated ICSI platform comprises both hardware and software components. The hardware includes an inverted microscope, stage movement control device, camera, fine adjustment device, laser objective lens, motor control device, injector, piezoelectric actuator, microtool holder, and 3D printed dish holder. The software components include AI / ML, software for operating the optics and microscope, and additional devices. The system can operate by executing a series of pre-programmed processes that can be started with a single command issued from a computer. Procedures can also be executed with multiple commands, each intended for a specific process within the procedure. The computer can be operated remotely by a person, and the manipulators and / or microscopes used throughout the process can be operated by robots. Imaging throughout the process can be visualized on a computer screen.

[0305] The IVF / ICSI platform can automate ICSI procedures, including sperm preparation and fixation, oocyte processing, zona pellucida excision, oocyte membrane disruption, sperm deposition, and oocyte release. The system can also provide control for microscope focusing, pipette pressure control, and stage and pipette movement.

[0306] After the automated ICSI procedure, an operator can remove the ICSI dish from the microscope stage, wash the oocytes manually using a pipette, and culture the oocytes in an incubator. Alternatively, these processes of removing the ICSI dish from the microscope stage, washing the oocytes manually using a pipette, and culturing the oocytes in an incubator can be performed automatically and robotically by the IVF / ICSI platform.

[0307] The disclosed system can provide a more consistent and reliable approach to ICSI. By automating these processes, the system can reduce variability, improve the success rate of ICSI procedures, and potentially increase the efficiency of ART.

[0308] In the context of ICSI, automation includes the use of robotic mechanisms and AI / ML assistance, as described herein, resulting in certain processes in microinfusion procedures that would normally require highly skilled operators being performed by intelligent robotic systems on the IVF / ICSI platform.

[0309] In some embodiments of the present invention, the IVF / ICSI platform can complete the fertilization of human eggs. Human eggs can be fertilized in vitro by injecting semen. Fertilization can be achieved by adding thousands of sperm to one or more eggs (referred to as conventional or standard IVF) in a dish, with the cumulus and corona radiata cells intact. Alternatively, fertilization can be achieved by microsurgical injection of a single sperm into a single egg from which all somatic cell coverings—cumulus and corona radiata cells—have been removed (referred to as intracytoplasmic sperm injection or ICSI). ICSI is a technique that involves microinjecting a single (immobilized) sperm directly into the cytoplasm of an oocyte to achieve fertilization. This procedure was primarily developed for the treatment of male factor infertility, where sperm may not be able to fertilize the egg "naturally," but is increasingly being applied for the treatment of non-male factor infertility / low fertility rates, as well as for vitrified / warmed eggs. Thus, ICSI can be the optimal treatment for essentially all patients seeking ART. To fertilize an egg, sperm must pass through the zona pellucida surrounding the egg, attach to the egg cell membrane, and enter the cytoplasm. In ICSI, the zona pellucida and egg cell membrane are bypassed, and sperm are delivered directly to the cytoplasm, where a series of fertilization and developmental events begin. Sperm are motile due to the vibration of their tails. This characteristic is important for achieving fertilization in vivo, as it allows sperm to traverse the female reproductive tract and reach the egg. Before injecting spermatocytes into the egg during ICSI, the motility of the cells can be stopped. This is called immobilization and can be achieved using mechanical pressure on the tail with an ICSI needle or laser shot. Lasers, including but not limited to 1460 nm, infrared solid-state diode, Class 1 lasers delivered through a 40x objective lens on an inverted microscope, can be used by IVF / ICSI platforms as part of carrying out ICSI and related processes. This laser can excise the zona pellucida without damaging the egg or embryo. Zona pellucida excision can be used for different purposes.In piezoelectric ICSI, piezoelectricity is used to facilitate the passage of the ICSI needle through the zona pellucida (by reducing its thickness), thus avoiding mechanical stress on the oocyte. A piezoelectric actuator is a transducer that converts electrical energy into mechanical displacement. Piezoelectric ICSI procedures can be modified compared to conventional ICSI. The piezoelectric drive unit is mounted in a microtool holder and can deliver pulses through the ICSI needle, traversing the zona pellucida, reaching the oocyte membrane, and pressing past the midpoint of the oocyte. The pulses can disrupt the membrane, allowing for the release of sperm into the cytoplasm.

[0310] In some embodiments, polyvinylpyrrolidone (PVP) is a high molecular weight molecule used in solution to slow sperm motility. This facilitates the capture, immobilization, and collection of sperm for injection into eggs, and prevents cell or cytoplasm from "adhering" to microscopy needles and pipettes.

[0311] In some embodiments, pipetting is a fundamental IVF laboratory technique and can be performed in a sterile manner without generating bubbles that can be destructive, lead to cell loss, or generate potentially infectious aerosols. Pipetting in the context of ICSI involves the aspiration, draining, and transfer of individual or multiple oocytes and bodily fluids from container to container. Types of pipettes that can be automatically and robotically used by IVF / ICSI platforms include, but are not limited to, 1) plastic capillaries with narrow holes (e.g., 175 μm inner diameter) for moving oocytes from dish to dish, 2) heated and stretched glass (Pasteur) pipettes to form narrow holes for moving oocytes from dish to dish, and 3) microsurgical pipettes or needles for holding oocytes in place and performing sperm injection.

[0312] In some embodiments, different types of pipettes can also be used automatically and robotically by an IVF / ICSI platform during ICSI, including, but not limited to, the following: 1) Eppendorf pipettes (e.g., used during dish preparation): featuring a piston and spring-loaded toothed cone, a single pathway and adjustable volume (in units of 1 to 1000 μL with a specific range), used in conjunction with an Eppendorf tip to aspirate and dispense precise (usually low) volumes; 2) Capillary tips: the tips are made from flexible medical-grade plastic to prevent scratching of plastic petri dishes. The tips are manufactured with different inner diameters ranging from 75 μm to 600 μm, with the most commonly used sizes being 155-200 μm for oocyte detachment and processing of oocytes and embryos, and 300 μm for blastocyst processing. 3) Microsurgical pipettes used in conjunction with fine-tuning devices: Microsurgical pipettes such as holding pipettes or ICSI needles / pipettes, and biopsy needles are ultrafine glass instruments manufactured with specific outer and inner diameters for holding or manipulating oocytes and embryos during microsurgical procedures with fine-tuning devices.

[0313] In some embodiments, the retaining pipette can have an outer diameter of 65 to 180 μm and an inner diameter of 15 to 30 μm.

[0314] In some embodiments, the ICSI needle (or pipette) has an inner diameter of 5-6 μm, a spike at the tip, and a 30-degree bend approximately 1 mm from the tip, allowing for use with a flat dish during microscopy.

[0315] In some embodiments, the IVF / ICSI platform can automate the ICSI process using robotic mechanisms, optics, and AI / ML (including machine vision) to perform ICSI in a fully automated manner. The entire ICSI procedure can be executed by a single command issued locally or remotely from a computer. For example, the command can initiate a series of pre-programmed ICSI processes, as described herein. The procedure can also be executed by multiple commands, each intended for a specific process within the procedure.

[0316] In some embodiments, the IVF / ICSI platform may use components in the execution of the ICSI process, including, but not limited to, the following: Inverted microscope Stage movement control device camera Stage heating and control device Fine adjustment device A range of 12,500 μm for each axis. Maximum speed 10,000μm / sec Laser objective lens Motor control device Injector (air) for holding the pipette Eppendorf injector for injection needles (oil) Piezoelectric actuator 3D printed dish holder (with direction markers) Stepping motor for micro-injectors Microtool holder Condenser Directional 3D printed dish holder

[0317] In some embodiments, the IVF / ICSI platform can use software to automatically operate AI / ML, optics, robotic mechanisms and microscopes, as well as other additional devices.

[0318] In some embodiments, examples of actions taken by the IVF / ICSI platform when performing ICSI and related processes include, but are not limited to, the following: 1. Calibrate the microtool and droplet position. 2. Place one or more oocytes on a prepared dish, and then place the dish on a heating stage for an inverted microscope, OCT or OCM, or other form of microscopy, including lensless microscopy. 3. The sequence of ICSI processes is initiated by issuing a single command from a computer, based at least partially on the AI / ML process, output, or result. Alternatively, individual commands can be issued manually or automatically for each process in the procedure. 4. The person operating the computer can do so remotely. The operator does not directly use a manipulator or microscope. Images can be visualized on a computer screen or through the integration of augmented reality (e.g., mixed virtual reality). 5. Commands can be presented as "buttons" on a user interface such as a computer screen (IDEM). For example, a set of commands may be used to instruct the IVF / ICSI platform during sperm processing, as follows: a. Prepare the needle (the injection needle can be lowered into the PVP droplet and "primed" with a predetermined volume of PVP, meaning that the PVP is drawn into the needle in preparation for sperm collection). b. Transfer to sperm droplet (The IVF / ICSI platform can automatically position the relevant cells and instruments in place for operations such as immobilization, filling, injection, infusion, or any other action. For example, based on initial calibration, the AI / ML / robot mechanism can recognize the location of the sperm droplet, automatically move the needle to that location, automatically select a 40x objective lens, and adjust the focus so that the sperm cells are in focus and clearly identifiable.) c. Immobilizing sperm (The AI / ML / robot mechanism can evaluate and score the motility of spermatocytes within the field of view, select the "best" sperm according to an algorithm, move the stage so that the selected sperm can be targeted by a laser for immobilization, and emit laser shots toward the tail to stop or slow down sperm motility, i.e., the sperm are immobilized. AI / ML-induced sperm immobilization can be based on real-time division intentionally incorporated to evaluate morphology and select the best position for striking or laser shots for immobilization.) d. Needle at the collection site (AI / ML / robot mechanism can determine the best position for sperm collection by the tail end) e. Sperm Filling (AI / ML / Robot Mechanisms can perform controlled sperm aspiration while evaluating the position of sperm being aspirated into the needle and leaving sperm in a predetermined position within the needle, at least in part on computer vision guided by AI / ML algorithms, e.g., control (e.g., pressure, fluid dynamics, or other calculations). f. Movement of the needle from the droplet (The needle can be lifted from the sperm / PVP droplet (e.g., it can be on the Z-plane and remain stationary in a predetermined position, such as for positioning for the next step (e.g., needle / instrument arrangement and ICSI)). 6. Another set of commands may be used to direct the IVF / ICSI platform, e.g., optics, associated equipment and stage, individually or in combination, for the purpose of positioning cells and appropriate equipment in the processing of oocytes, as follows: a. Move to the egg (The stage can be moved so that the egg moves to, for example, a 4x magnification field of view, and the AI / ML / robot mechanism can identify and position the egg, change the magnification to, for example, 40x, and adjust the focus so that the zona pellucida is clear.) b. Holding the oocyte (The oocyte can be held in place through a physical barrier (e.g., the bottom of a dish, a designed wall, or a blunt concave pipette) or by a holding pipette. In one example, the holding pipette can be lowered into the droplet, and the AI / ML / robot mechanism can determine the best area for holding the oocyte and bring the holding pipette to that position. The AI / ML / robot mechanism can apply negative pressure to the pipette to collect and hold the oocyte with the holding pipette.) c. Open zona pellucida (The AI / ML / robot mechanisms and optics described herein can determine the location and amount of zona pellucida required for excision, and where zona pellucida excision should be performed, move the oocyte to the target location, and generate laser shots to excise a predetermined area and depth of the zona pellucida. The AI / ML / robot mechanisms can assess the thickness of the zona pellucida and determine the intensity and radius required for one or more shots to effectively excise it, and / or determine the number of laser shots required to excise the appropriate portion of the zona pellucida to facilitate the entry of the ICSI needle through the zona pellucida without distorting the oocyte.) d. Define the injection route (AI / ML / robot mechanisms can determine the best injection route, including the initial (outside the zona pellucida) and final (intracytoplasmic) needle positions, based, for example, on planar images (real time) and / or through pre-loaded information based on 3D modeling and / or imaging of the oocyte. OCT can help identify the equatorial plane of the oocyte and program the needle's path from the entry site to where the needle tip should stop after membrane perforation.) e. Sperm at the tip (AI / ML / robot mechanisms can apply a combination of injection / aspiration (or positive and negative pressure) to position sperm at the desired location. For example, if desired, positive pressure within the ICSI needle can be used to push sperm to the tip of the needle in preparation for injection into the cytoplasm, with the aim of reducing the volume of PVP (or other exogenous medium) injected into the egg cytoplasm.) f. Penetration of the oocyte (The AI / ML / robot mechanism can move the needle forward into the oocyte at a predetermined controlled speed and stop when it reaches the end of the path. This can be predetermined by other means, including the use of conventional microscopy, OCT, OCM, or 3D reconstruction of the oocyte, and may or may not be adjusted in real time based on the AI / ML modeling characteristics.) g. Disrupting the egg cell membrane (The egg cell membrane can be disrupted using piezoelectric pulses. AI / ML / robot mechanisms can be used to determine whether piezoelectric pulses are necessary to disrupt the membrane and how many pulses are required.) h. Sperm deposition (AI / ML / robot mechanisms can control and apply positive pressure within the ICSI needle to deposit sperm into the egg.) i. The egg is removed (The AI / ML / robot mechanism can determine that the sperm is outside the needle and move the needle out of the egg.) j. Egg release (The AI / ML / robot mechanism can control and apply positive pressure within the holding pipette until the egg is released into the droplet.)

[0319] In some embodiments, other commands, including, for example, "piezoelectric pulse" and "emit laser," may be available and, if necessary, intended for human interaction / intervention with the IVF / ICSI platform.

[0320] For example, the user interface and controls of the IVF / ICSI platform can be made available for magnification, microscope focus, incremental negative and positive pressure control in the holding pipette, and positioning of the holding pipette. Controls can also be made available for stage movement, incremental negative and positive pressure control in the ICSI needle, and positioning of the ICSI needle.

[0321] The IVF / ICSI platform can automatically remove the ICSI dish from the microscope stage, wash the oocytes using automated pipetting, and place the oocytes in a culture dish in an incubator. Alternatively, the removal of the ICSI dish can be performed manually by an operator.

[0322] Table 4 shows a comparison between conventional ICSI and piezoelectric ICSI. (P) indicates a process specifically for piezoelectric-assisted ICSI (including ICSIA).

[0323] [Table 4]

[0324] In some embodiments of the present invention, the IVF / ICSI platform described herein may include a fertilization module. The fertilization module may be equipped with an inverted microscope having an electric stage on which a specimen is placed, light is projected from above the specimen, and the microscope lens is below the specimen. In some embodiments, the electric stage may operate independently or in conjunction with the electric lens. A fine adjustment device may be used with the fertilization module, and the fine adjustment device has motors that allow movement in three axes, namely X, Y, and Z.

[0325] In some embodiments, the insemination module can receive an ICSI dish containing a specimen and use imaging and AI / ML processes as described herein, together with an inverted microscope and fine-tuning device. In some embodiments, the specimen may contain multiple droplets, each containing at least one oocyte or sperm from the oocyte and specimen prepared by the semen preparation module, respectively, as described herein. Sperm can be treated with PVP to slow sperm motility. In some embodiments, the ICSI dish can be covered with oil to minimize evaporation. Imaging can facilitate the positioning of the specimen relative to the insemination module's equipment, such as robotic devices, illumination, microscopy, and other devices, enabling fully automated control of the specimen and standardization of the robotic process. Figure 41 shows an exemplary embodiment of an adapter for using a ring light with a standard needle holder that can be used within the insemination module (4100). Figure 42 shows an adapter for using an XY positioning system with a microscope neck that can be used within the insemination module (4200). Figure 43 shows a mount for a cell sorter and / or cell harvester for use within the fertilization module (4300). In some embodiments, the imaging and AI / ML processes of the fertilization module can instruct the robotic mechanism of the fertilization module to position droplets in the same location or near the same location during each fertilization procedure. This reduces the time required to position and utilize the eggs and sperm within the droplets.

[0326] In some embodiments, the insemination / infusion module may use an air injector having at least one motor for controlling suction, which can be remotely controlled as part of the robotic mechanism of the insemination module. Figures 44–48 show devices that can be used to rotate one or more knobs of a fluid aspirator that can be used within the insemination module (4400–4800).

[0327] In some embodiments, the fertilization module may use a game engine or multiple game engines to construct computer vision algorithms to assist in identification within microscopic imaging, for example, detecting oocytes, detecting the tip of a needle, detecting the tip of a holding pipette, detecting sperm inside the needle, detecting oocytes at different magnifications, or detecting any other target or reference state. In some embodiments, the fertilization module may use a user interface to robotically control one or more microscopes, fine-tuning devices, etc. In one example, the fertilization module can evaluate a sample droplet that is known to contain oocytes, but whose exact location is unknown. In some embodiments, the robotic mechanism of the fertilization module can then use a needle to extract oocytes from the droplet held in an ICSI dish. The fertilization module may be able to automatically adjust the magnification used by the microscope for the purpose of imaging the droplet in which the oocytes are located. Using the fertilization module's user interface, the user can initiate commands by selecting to press buttons within the user interface, and the stage containing the sample can be moved to a given position. Subsequently, the imaging and AI / ML processes described herein can activate the image of the oocyte and begin searching. If an oocyte is detected, the fertilization module can initiate a routine to move the stage and center the oocyte. Using the imaging and AI / ML of the fertilization module, the central position of the oocyte can be determined, and the stage can be robotically maneuvered to center the oocyte. Once the oocyte is centered, the microscope magnification can be automatically changed again (e.g., to 40x magnification), and the center of the oocyte can be used as a focusing target for the imaging and AI / ML processes. The algorithm can detect the oocyte's equator and then use it to calculate the position where the needle can enter the sample and / or where the pipette should engage with and hold the oocyte.

[0328] In some embodiments, a machine / computer vision system on the fertilization module can detect sperm and evaluate individual sperm using multiple cameras, possibly positioned at different angles, to determine which are the best candidates for use in embryo transfer. Following this identification, selected sperm can be immobilized by automatically using a laser, for example, to block tail function or to apply one or more short piezoelectric actuator pulses. In some embodiments, the sperm module imaging and AI / ML process can, for example, use segmentation to determine the best location along the sperm tail and label specific areas of the sperm, such as the midpiece, the middle of the tail, or the tip of the tail, to direct laser or piezoelectric cutting or blocking, helping to minimize the possibility of DNA damage. Following sperm tail immobilization, the fertilization module imaging and AI / ML process can be used to detect a lack of desired sperm motility. If a sperm is not sufficiently immobile, the procedure of laser targeting the sperm's location and tail can be repeated for other candidate sperm, and the imaging and AI / ML process of the fertilization module can repeat the process for multiple sperm candidates until it is confirmed that the candidate sperm is sufficiently immobile. In some embodiments, the imaging and AI / ML process of the fertilization module can detect the positioning of the sperm, and as a result, the robotic mechanism of the fertilization module can be instructed to orient the robotic device to the optimal position for sperm acquisition, such as by using a needle with the robot to capture the candidate sperm.

[0329] In some embodiments, once sperm are obtained, the fertilization module can activate aspiration. Using the control system of the fertilization module, it is possible to determine how to aspirate and track the sperm and how to do so by using imaging and AI / ML processes as described herein.

[0330] In some embodiments, the fine-tuning device can hold the oocyte and inject the sperm. Figure 49 shows an adapter that can be used to connect a linear axis to an XY micropositioning device that can be used within the fertilization module (4900). The oocyte can be held within the fertilization module using motor-driven suction with a suction device, and a needle can be positioned at the entry point. Sperm can be placed in the tip of the needle, the needle can be inserted into the oocyte, the membrane can be broken, and the needle can be removed once the sperm has been released from the needle. In one exemplary embodiment, the fertilization module can use laser or piezoelectric-assisted ICSI to open a pathway within the zona pellucida, through which the sperm can penetrate the oocyte and deposit within the oocyte, thereby minimizing the risk of deformation of the oocyte. Imaging and AI / ML processes of the fertilization module can be used after injection to confirm that the oocyte was not damaged or killed during the fertilization process.

[0331]

[0200] In some embodiments, once an egg is fertilized, the fertilization module can prepare the embryo for incubation. Once the embryo is incubated, the IVF platform can continue to culture the embryo for a specified period (e.g., 5-7 days). In some embodiments, imaging and AI / ML processes of the IVF platform can be used to periodically or continuously monitor cell division occurring within the embryo and algorithmically determine the embryo's health, developmental stage, viability (or other predictors). In some embodiments of the present invention, the IVF / ICSI platform described herein can autonomously select, immobilize, and process single or multiple spermatocytes and automatically prepare these cells for injection into oocytes. In one embodiment, the IVF / ICSI platform can simultaneously or nearly simultaneously identify, select, immobilize, and aspirate motile sperm into a microneedle. For example, an IVF / ICSI platform can perform the following processes, including, but not limited to, those listed below, which can be initiated digitally as separate functions, or as a set of functions digitally using a single command. (1) Sperm identification (SID) (2) Motorized staging and processing (3)Auto focus (4) Tracking sperm during processing (5) Identification and tracking (fractionation) of the sperm tail (6) Immobilization of the sperm tail by (a) laser, (b) machine, or (c) piezoelectric (7) Descending and automatic positioning of the sperm injection microneedle (8) Filling and positioning of sperm within the needle

[0332] In some embodiments of this disclosure, the fertilization module may include fertilization components, systems, and processes. In some embodiments of this disclosure, the fertilization module may include AI / ML for automated detection, identification, and classification. In some embodiments of this disclosure, the fertilization module may include AI / ML for automated measurement and testing. In some embodiments of this disclosure, the fertilization module may include AI / ML for optimization. In some embodiments of this disclosure, the fertilization module may include AI / ML for prediction. In some embodiments of this disclosure, the fertilization module may include AI / ML for selection / rating. In some embodiments of this disclosure, the fertilization module may include AI / ML for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the fertilization module may include AI / ML for system configuration and control. In some embodiments of this disclosure, the fertilization module may include fully autonomous AI / ML. In some embodiments of this disclosure, the fertilization module may include optical and machine vision components, systems, and processes. In some embodiments of this disclosure, the fertilization module may include robotic processing systems and processes. In some embodiments of this disclosure, the fertilization module may include sensor components, systems, and processes. In some embodiments of this disclosure, the fertilization module may include specimen management components, systems, and processes.

[0333] In some embodiments of the present invention, the IVF / ICSI platform described herein can use robotic control to mimic human pipetting for oocyte, sperm, and embryo transfer without microfluidics. In one example, the IVF / ICSI platform can perform the following processes, including, but not limited to, the following: (1) Autonomous semen pipetting (2) Viscosity test (3) For example, stratification of sperm / culture medium using a two-layer or three-layer (multilayer) system. (4) Viscosity reduction by pipetting and / or ultrasound (5) Optical detection of "swim-up" efficiency by comparing the transparency of the upper layer before and after a certain period of time. (6) Automatically position the capillary pipette tool vertically, horizontally, or at an angle for pipetting the oocyte. Remove the cumulus oophorus and corona radiata cells using the automation of the wide-mouth capillary pipette while the oocyte is in hyaluronidase. (7) Visualization of the position of the oocyte at the tip of the pipette and within the pipette is not necessary for the complete movement of the oocyte. (8) The same applies to embryos at any developmental stage to which the process (e.g., processes 6 and 7 above) is applied.

[0334] In some embodiments, the IVF / ICSI platform may use a control system, such as vision-based robot control (visual servo), to control the movement of the robotic apparatus, including but not limited to the following: • Move the stage to the specified X, Y Cartesian position. • Start / stop / reverse attraction to a specified increment • Laser sperm immobilization • Precise positioning of the needle • Precise positioning of the retainer • Intraneous sperm positioning

[0335] For example, the control system of an IVF / ICSI platform can be used with a visual servo to precisely position sperm tails at designated spots within the laser path.

[0336] The purpose of the visual servo is to position the sperm tail within the laser path, as indicated by the laser spot (5000), i.e., to minimize error e(t). e(t)=s(t)-s* (3.1)

[0337] Here, s(t) = [x 尾部 , y尾部 T is the obtained sperm tail position in the image frame X i O i Y i (FIG. 50), and s* is the position of the laser spot within the image frame, which is a constant. FIG. 51 shows an exemplary system control diagram (5100). FIG. 52 shows a simplified system control diagram for laser sperm ablation (5200).

[0338] In one example, the control system of the IVF / ICSI platform can be used for visual servoing to position a needle, pipette, or other device. FIG. 53 shows a simplified system control diagram for precise needle positioning (5300). For example, using visual servoing, the needle tip can be positioned inside a specified location, that is, the error e(t) can be minimized. e(t)=n(t)-D * (3.1)

[0339] Here, n(t)=[x needle, Y needle]T is the obtained needle tip position in the image frame X i O i Y i and D * is the specified spot in the image frame, which is a constant. needle ti D *

[0340] In one example, the control system of the IVF / ICSI platform can be used for visual servoing to position a holder, microtool holder, or other device. FIG. 54 shows a simplified system control diagram for precise holder positioning (5400). For example, using visual servoing, the holder tip can be positioned at a specified location, that is, the error e(t) can be minimized. e(t)=n(t)-D* (3.1)

[0341] Here, n(t)=[x holder, Y holder]T is the image frame X i O i Y iThis is the needle tip position obtained in the image, and D* is a constant, specified spot within the image frame.

[0342] Figure 55 shows a simplified system control diagram for the precise positioning of sperm inside the needle (5500).

[0343] Figure 56 shows a simplified system control diagram for a typical cell aspiration configuration that can be used by an IVF / ICSI platform (5600).

[0344] Figure 57 shows a simplified diagram of a model-based adaptive control system that can be used by the IVF / ICSI platform (5700).

[0345] For example, the control system of the IVF / ICSI platform can be used to control stage movement, and the stage position can be set to a given X,Y coordinate using pixels.

[0346] For example, the control system of an IVF / ICSI platform can be used to perform a series of operations, including, but not limited to, the processes in the following ICSI-related sequences: Locate sperm droplets and oocytes. Move to sperm droplets Select and immobilize the chosen sperm. The selected sperm are loaded into the needle. Move to oocyte droplets Retaining oocytes Penetrating the oocyte Deposition and confirmation that sperm have entered the cytoplasm. SI End

[0347] Figure 58 shows a simplified workflow for locating sperm droplets and oocytes that can be used by the IVF / ICSI platform (5800).

[0348] For example, the process of "locating sperm droplets and oocytes" may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Set the "infocusdrop" flag to false. 2. Set the "End sweep" flag to false. 3. Use a simple stage control device to move the stage to its initial position. 4. While "End sweep" is false 1. If "infocusdrop" is false 1. Use droplet edge detection to determine if the current location is the edge of a droplet. 1. If it is possible that it is the edge of a droplet, adjust the focus and use a simple focus control device to obtain a clear image of the droplet edge and determine whether the focus is on the droplet edge. 2. If the edge of the droplet is visible, register the focal motor position for sperm (FS). 3. Set "infocusdrop" to true. 2. If "infocusdrop" is true 1. Use "Determine if sperm drop" to determine if the current location is a sperm drop. 3. Use the "Detect Oocytes" function on the screen to determine if there is a possibility of it being an oocyte. 1. If there is a possibility of an oocyte on the screen, adjust the focus and use "Determine ZP at Focus" to make the ZP clearly visible. 2. Register the focus motor position of ZP (FZ) 4. Register the location type on the map (None on screen, Sperm droplet, Oocyte) 5. Move to the next location. 6. When the final location is reached, set "End sweep" to true. 7. Calculate the output from the map using connected components. • Central position of sperm droplet (SDP) • Central position of the oocyte (OCP)

[0349] For example, the process of “moving to a sperm droplet” may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Use a simple focus control device to set the focus to FS. 2. Use a simple stage control device to move the center of the stage to the "SDC" coordinates.

[0350] For example, the process of “selecting and immobilizing selected sperm” may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Use a simplified focus control device to set the focus to position FS. 2. Start tracking motile sperm. 3. After 10 seconds, 4. Grade the sperm on the screen by motility and use sperm grading to identify sperm A. 5. Determine the position of the tail (STP) of sperm A at this moment. 6. Determine the distance DLS between the STP and the laser target (LT). 7. Between DLS and TOL_LS 1. Perform iterations of advanced stage control systems to reduce the difference between DLS and laser target LT. 2. Determine the STP (Sequence Position of Particle A) of the tail of sperm A at this moment. 3. Determine the distance DLS between the STP and the laser target LT. 8. Use "Emit Laser" to fire a laser. 9. Use "Detect selected non-motile sperm" to confirm that the selected sperm have been immobilized. 10. Register the center of the sperm head position (SHP).

[0351] For example, the process of "filling a needle with selected sperm" may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Use "Detect selected non-motile sperm" to determine the sperm retrieval location (SPP) from the sperm head location (SHP). 2. Use a simple needle control device to move the needle to the SPP. 3. Using the simple needle and the "detect whether the needle is in focus" function, lower the needle until it is in focus. 4. Detect the needle tip NT. 5. Determine the distance DNT_SPP between the needle tip NT and the sperm collection site SPP. 6. Between DNT_SPP and TOL_PS 1. Reduce DNT_SPP by performing an iteration of an advanced needle control device. 2. Determine DNT_SPP 7. Determine the target intra-needle sperm location (TNSP). 8. Determine the distance D_STN between the sperm head positions SHP and TNSP. 9. Between D_STN > TOL_SIN 1. Perform repeated, advanced suction to reduce D_STN. 2. Determine the position of the selected sperm head (SHP) at this moment using "Detect selected non-motile sperm" and "Detect sperm position inside the needle". 1. The outside of the needle 2. Inside the needle 3. Determine D_STN 10. Using the simple needle and the "Detect whether the needle is in focus" function, raise the needle until it is out of focus.

[0352] For example, the process of “transferring to an oocyte droplet” may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Use a simple stage control device to move the center of the stage to the "OCP" coordinates. 2. Use a simple focus control device to set the focus to FZ.

[0353] For example, the process of "holding an oocyte" may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Determine the oocyte holding posture using "Detect the center of the oocyte". 2. Use a simple stage control device to move the stage to the oocyte-holding position. 3. Determine the initial pipette holding position HPI and the final pipette holding position HPF. 4. Transfer the retained pipette to the HPI using the simple retaining pipette control device. 5. Using the simple pipette holding control device and the "detection of whether the pipette is in focus" function, lower the pipette until it is in focus. 6. Detect the HPT at the tip of the retaining pipette. 7. Determine the distance DHPT_HP between the holding pipette tip HPT and the holding position HPI. 8. Between DHPT_HP and TOL_HPP 1. Reduce DHPT_HP by performing iteratives on advanced pipette retention control devices. 2. Determine DHPT_HP 9. Repeat from step 6 for HPF. 10. Activate aspiration using a simple pipette aspiration control device. 11. Using "Detect the center of the oocyte" and "Detect the tip of the holding pipette," move the holding pipette up and down with the simple holding pipette control device while tracking the position of the tip of the holding pipette (HPT) and the center of the oocyte (COO). 12. If the vertical coordinates of the tip of the holding pipette (HPT) and the center of the oocyte COO move in the same direction, stop at the original vertical coordinate.

[0354] For example, the process of "penetrating the oocyte" may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Use a simple needle control device to move the needle to the IPP. 2. Using the simple needle control device and the "detection of whether the needle is in focus" function, lower the needle until it is in focus. 3. Use "Detect needle tip" to detect the needle tip (NT). 4. Use an advanced sperm aspiration control device to place the sperm to be injected into the needle tip. 5. Determine the distance DNT_IPP between the needle tip NT and the initial penetration position IPP. 6. Between DNT_IPP and TOL_IPP 1. Reduce DNT_IPP by performing iterative advanced needle control. 2. Determine DNT_IPP 7. Determine the final penetration point (FPP) using "Detect the center of the oocyte" and "Detect the cytoplasmic boundary". 8. Calculate the distance D_FPP between the needle tip NT and the final penetration position FPP. 9. Between D_FPP > TOL_IPP 1. Reduce D_FPP by performing iterative advanced needle control. 2. Determine D_FPP

[0355] For example, the process of "depositing and confirming that sperm have entered the cytoplasm" may include the control system instructing an automated robotic system to initiate actions, including, but not limited to, the following: 1. Confirm that the sperm are deposited and located within the cytoplasm. 2. Use "Detect needle tip" to detect the needle tip (NT). 3. Use "Detect sperm location inside the needle" to detect the sperm location SHP inside the needle. 4. Determine the target intra-needle sperm location (TNSP). 5. Determine the distance D_HSTN between the sperm head positions SHP and TNSP. 6. Between D_HSTN > TOL_TNSP 1. Perform iteratives of advanced suction control to reduce D_HSTN. 2. Use "Detect sperm location inside the needle" to determine the immediate location of the selected sperm head (SHP) inside the needle. 3. Determine D_HSTN 7. Determine the sperm release site (SRP). 8. Determine the distance D_SRP between the sperm head position SHP and SRP. 9. Between D_SRP > TOL_SRP 1. Reduce D_SRP by performing iteratives with advanced suction control devices. 2. Use "Detect sperm location inside the needle" to determine the immediate location of the selected sperm head (SHP) inside the needle. 3. Determine D_SRP 10. Use "Detect sperm in the cytoplasm" to confirm sperm in the cytoplasm. 11. Use "Detect sperm location inside needle" to confirm that there are no sperm inside the injection needle.

[0356] For example, the process of "terminating ISCI" may include the control system instructing the automated robotic system to begin an action that includes, but is not limited to, the following: 1. Use a simple needle control device to move the needle to IPP+delta. 2. Apply negative pressure to the holding pipette using a simple holding pipette aspiration control device. 3. Use "Detect Needle Tip" and "Detect Outside ZP" to determine the initial oocyte pressure position (OPIP) using the central and outer contours of the oocyte zp. 4. Use a simple needle control device to move the needle to the OPIP. 5. Use "Detect Center" to detect the central IOC of early oocytes. 6. Use a simple needle control device to move the needle to the stable central IOC of the oocyte coordinate system. 7. When the central FOC of the final oocyte is detected, it should be different from the IOC. 8. Using the simple needle control device and the "detection of whether the needle is in focus" function, raise the needle until it is out of focus. 9. Using the simple pipette holding control device and the "detection of whether the pipette is in focus" function, raise the pipette until it is out of focus.

[0357] In some embodiments, the IVF / ICSI platform may use a control system, such as vision-based robot control (vision servo), to control the movement of the robotic apparatus, including, but not limited to, needle control device commands such as the following: Move to position (X, Y, Z, velocity x, velocity y, velocity z). ·Stop ·Movement X (number of processes x, speed x) ·Movement Y (number of processes x, speed x) ·Movement Z (number of processes x, speed x)

[0358] In some embodiments, the IVF / ICSI platform may use a control system, such as vision-based robot control (visual servo), to control the movement of a robotic device, including, but not limited to, pipette control device commands such as, but not limited to, the following. Move to position (X, Y, Z, velocity x, velocity y, velocity z). ·Stop ·Movement X (number of processes x, speed x) ·Movement Y (number of processes x, speed x) ·Movement Z (number of processes x, speed x)

[0359] In some embodiments, the IVF / ICSI platform may use a control system, such as vision-based robot control (visual servo), to control the movement of the robotic device, including, but not limited to, pipette aspiration control device commands such as the following: • Movement (number of steps, speed) (where, for example, speed can be negative) ·Stop

[0360] In some embodiments of the present invention, the IVF platform described herein can monitor the development of multiple embryos within an incubation module. One challenge with conventional box-type incubators is that when the incubator is opened, room air enters and can adversely affect the temperature inside the incubator. Also, gases can escape and affect the pH of the culture medium. When the incubator door is closed, time is required to restore the specified gas level and temperature, all of which can be harmful to developing embryos. Partly for this reason, the industry has shifted significantly to flat tabletop incubators with smaller chambers having smaller internal volumes. The smaller the air volume, the faster the recovery rate can be in terms of gas and temperature recovery. In some embodiments, the incubation module of the IVF platform can use general-purpose culture media for processing developing embryos, thereby allowing the incubator holding the embryos to remain closed or not be opened very frequently.

[0361] In some embodiments, the incubation module of the IVF platform may use the imaging and AI / ML processes described herein as part of the incubation stage. For example, images of the embryo may be taken over a selected period of time, continuously, or in some combination of fixed and continuous imaging, at regular or fixed intervals. Imaging may include still images, images over time, simulated images, images with inferred characteristics, and / or animated sequences of moving or still images such as videos. The imaging and AI / ML processes of the incubation module can automatically assist in the detection of relevant events in cell and embryo division, for example, when two pronuclear "circles" are observed, confirming normal fertilization. In some embodiments, the incubation system of the incubation module of the IVF platform may integrate the imaging system and camera system within the incubator. The incubator may include a dual complex in which the incubator has a microscope system integrated therein, together with a robotic mechanism. A dual complex, for example, can have two controlled environments within the entire incubator: one where embryos can be present during development, and a second that can be used to automatically image individual embryos or groups of embryos. In one example, a robotic mechanism, at least inside the incubator, can automatically select a dish containing embryos or multiple embryos and automatically transfer the dish from a first compartment used for embryo development to a second compartment within the same incubator where microscopy and imaging can be performed. After imaging, the robotic mechanism can automatically return the dish containing embryos or multiple embryos from the second compartment to the first compartment for further embryonic development.

[0362] In some embodiments, the IVF / ICSI platform can automatically classify multiple embryos based at least in part on the imaging and AI / ML processes of the incubation module. For example, the Embryo Grading Intelligent Classification Assistant (ERICA) described herein can be used by the incubation module to classify embryos, for example, based on the developmental stage and / or quality of each embryo. For example, the imaging and AI / ML processes of the incubation module can capture multiple planar images, cross-sections, etc., of each embryo and analyze these images to determine the presence of appropriate cell divisions, the rate of division, the nature of division, or any other biological markers of interest. The analysis may include quantitative analysis, including but not limited to the rate of cell division, as well as qualitative analysis, for example, by determining whether the divisions occurring are symmetrical and / or the shape of the divisions, the ratio of fluid area to cell mass, the area of ​​the cell mass, etc.

[0363] In some embodiments, embryos can be identified, tagged, and stored with biomarked identity using imaging, at least in part, based on imaging and AI / ML processes in the incubation module of an IVF / ICSI platform, so that the IVF / ICSI platform can associate a given embryo with the oocyte and sperm from which it originates, and so that embryos that have entered cryopreservation can be identified and stored with biomarked identity using imaging so that their biomarks can be confirmed and thus their identity can be confirmed upon retrieval and thawing from storage.

[0364] In some embodiments, the incubation module of an IVF / ICSI platform can provide clinical embryologists and / or male veterinarians with an automated, robotic testing platform for conducting clinical trials, observations, and other types of studies, assisting in the determination of the impact of environmental conditions related to incubation and vitrification on embryogenesis. The incubation module's ability to automatically monitor, image, measure, and predict aspects of embryogenesis can enable rapid collection of data on embryogenesis, which can be used to quickly optimize the incubation module's imaging and AI / ML processes without human intervention, and can inform external parties of aspects of IVF procedures that would benefit them. In one example, such “big data” projects can yield valuable insights that are currently financially unfeasible when human intervention and / or the use of human clinical embryologists and / or male veterinarians are required, including, to name just a few, measuring the impact of CO2 exposure and timing on blastocyst formation rates. In another example, such big data can yield other insights related to the application and timing of compounds to support and facilitate blastocyst formation. For example, individualizing the microfluidic fluid for embryos.

[0365] In one example, the ERICA system described herein can score and rank embryos based at least in part on imaging performed by the incubation module of the IVF / ICSI platform. The ERICA system can use still images, sequential imaging, simulated images, images with inferred characteristics, video and / or three-dimensional imaging. In some embodiments, the three-dimensional imaging used in the ERICA system can be based on real images automatically captured by the camera of the incubation module and / or on inferred, predicted, modeled or other non-real images that can be used to construct a three-dimensional model including, but not limited to, a digital twin of an embryo or multiple embryos. In some embodiments, a digital twin of a developing embryo can be advanced through a simulated life cycle and / or a predicted life cycle, at least in part on the AI / ML processing capabilities of the incubation module which can utilize the big data described herein. This simulated and / or predicted life cycle using digital twins can improve the accuracy and selection of embryos or groups of embryos that have the highest probability of success in achieving a desired clinical state, such as a specific developmental stage, the timing of reaching a particular developmental stage, or any other criterion. This predictive capability can improve the distinction between embryos suitable for cryopreservation and those that are not.

[0366] In some embodiments, the IVF / ICSI platform may include an intelligent layer that may include a digital twin system, which may include a set of components, processes, services, interfaces, and other elements for developing and deploying digital twin capabilities for visualization of various IVF existence states, properties, processes, methods, environments, and applications, as well as for coordinated intelligence (including artificial intelligence, analytical and other capabilities) and other services and capabilities enabled or facilitated by the digital twin. Without limit, the digital twin may be used and / or applied to each of the processes managed, controlled or mediated by each of the set of applications and processes of the IVF / ICSI platform described herein.

[0367] In some embodiments, the digital twin can leverage the presence of multiple applications within the IVF / ICSI platform, resulting in a pair of applications sharing data sources and other inputs collected for the operation and processes of the IVF / ICSI platform, and sharing outputs, events, state information, and results, which together can provide a much richer environment for enhancing the content of the digital twin, including the use of artificial intelligence (including any of the various expert systems, artificial intelligence systems, neural networks, supervised learning systems, machine learning systems, deep learning systems, and other systems described throughout this disclosure).

[0368] In some embodiments, the digital twin can be used in connection with a sharing or convergence process between various pairs of applications of the IVF / ICSI platform, such as components and modules of the IVF / ICSI platform described herein. In some embodiments, the convergence process may include a shared data structure for multiple applications that can be connected to the digital twin, such that the digital twin is updated accordingly.

[0369] For example, the ERICA system described herein can score and rate embryos based at least in part on imaging performed by the incubation module of an IVF / ICSI platform. Using still images, sequential imaging such as images over time, simulated images, images with inferred properties, videos, and / or three-dimensional imaging, the ERICA system can determine, for example, the amount of lipids in an embryo and predict the impact of the detected or predicted amount of lipids present on relevant clinical criteria, including, but not limited to, the probability of successful embryo transfer resulting in pregnancy.

[0370] In some embodiments, the dish, receptor, or container in which the embryos are held may be treated with a transparent oil, cellulose-based, or graphite-based film to reduce the rate of gas exchange and pH changes, and to provide an additional safety layer when moving cells to, from, and within the incubator of the incubation module.

[0371] In some embodiments, the incubation module's imaging and AI / ML processes may include a sperm-oocyte feature integration function (SOFI) that can be used to classify, score, and grade seemingly equivalent healthy embryos, at least in part, based on the measured or predicted quality of the sperm (or oocyte) from which the embryos originate. In some embodiments, the SOFI may function as a standalone algorithm or be incorporated to improve accuracy and add explainability to embryo grading and selection algorithms.

[0372] In some embodiments, imaging and AI / ML processes in the incubation module can be used to determine the optimal timing for cryopreserving embryos.

[0373] In some embodiments, the incubation module can reduce the amount of cryoprotectant required to process the embryos before cryopreservation, reduce or eliminate the vaporization step of entry into cryopreservation, accelerate the entry of the embryos into cryopreservation, and perform vitrification by consistently changing the angle of entry of the embryo container in a manner that minimizes the amount of air bubbles that may form around the container. In some embodiments, the vitrification performed by the incubation module can be performed automatically and robotically by the IVF / ICSI platform.

[0374] In some embodiments of this disclosure, the incubation module may include incubation components, systems, and processes. In some embodiments of this disclosure, the incubation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the incubation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the incubation module may include AI / ML for automated measurement and testing. In some embodiments of this disclosure, the incubation module may include AI / ML for optimization. In some embodiments of this disclosure, the incubation module may include AI / ML for prediction. In some embodiments of this disclosure, the incubation module may include AI / ML for selection / rating. In some embodiments of this disclosure, the incubation module may include AI / ML for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the incubation module may include AI / ML for system configuration and control. In some embodiments of this disclosure, the incubation module may include fully autonomous AI / ML. In some embodiments of this disclosure, the incubation module may include robotic processing systems and processes. In some embodiments of this disclosure, the incubation module may include advanced microscopy systems and components.

[0375] In some embodiments of the present invention, the IVF / ICSI platform described herein may include a vitrification module that provides equilibration and vitrification solutions before placing oocytes and / or embryos in or on a freezing device. As used herein, “automation” of vitrification means a vitrification process that includes the robotic mechanisms, optical systems and / or ML / AI / ML assistance described herein, so that processes in a vitrification procedure that normally require an operator can be performed by the intelligent robotic system of the IVF / ICSI platform.

[0376] Conventional vitrification procedures performed by operators generally follow the processes listed in Table 5 below.

[0377] [Table 5] TIFF2026524888000007.tif240170

[0378] In some embodiments, human germ cells and tissues can be stored indefinitely at cryogenic temperatures for future clinical or research use. When cells are cryopreserved, their cellular function ceases. To cryopreserve eggs and embryos, cryoprotective solutions (colloquially referred to as “antifreezes”) can be used to prevent cell death. Currently, a mixture of dimethyl sulfoxide (DMSO) and ethylene glycol (EG) can be used as a “permeable” (i.e., diffuses into cells) cryoprotective agent in combination with trehalose (or sucrose) as a “non-permeable” cryoprotective agent (i.e., remains outside the cell). The solution also contains proteins and macromolecules such as Ficol and hydroxypropylcellulose (HPC). Cryopreserved human germ cells and tissues can be stored at ultra-low temperatures (e.g., -196°C) to maintain their cryopreserved state. Cryopreservation can be carried out in dedicated tanks and relies on liquid nitrogen (LN2) or nitrogen vapor. Water is the main component of cells, and its coagulation (ice crystal formation) during cryopreservation can be controlled to avoid damage to organelles and membranes. Vitrification (as opposed to slow freezing) is a cryopreservation method that avoids ice crystal formation by changing the liquid to a glassy state using a high concentration of cryoprotectant, an ultra-rapid cooling rate (from room temperature to LN2), and suspension of cells in a small amount of cryoprotectant (<1 μL). Human oocytes and embryos can be efficiently and successfully cryopreserved using vitrification with a viability of >90%. Immediate viability means intact membranes (no thawing) in the case of oocytes, more than 50% of cells surviving in cleavage-stage embryos, and the majority of cells remaining intact in blastocysts. Ultimately, viability means retaining viability (e.g., fertilization ability in the case of oocytes, and continued development in the case of embryos).

[0379] In some embodiments, an equilibration solution (ES, non-vitrified) can be used. In one example, exposure to the ES allows cryoprotective agents (CPs, osmotic pressure approximately 2400) to be introduced into cells (osmotic pressure approximately 300). Intracellular water can leach out due to this difference between intracellular and extracellular osmotic pressure. Simultaneously, CPs can penetrate the cell membrane and enter the cell. As intracellular and extracellular osmotic pressures equilibrate, these reactions can occur simultaneously, but the rate at which water leachs out of the cell is slightly faster than the rate at which CPs enter the cell. Therefore, the oocyte / embryo first contracts, and then the water is replaced by CPs, allowing it to return to its original volume. The successful completion of equilibration can be indicated by the complete recovery of the cell volume.

[0380] In some embodiments, vitrification solutions 1 and 2 (VS1 and VS2) can be used. The purpose of VS1 is to replace all intracellular ES cells with VS. This process is completed when the intracellular osmolality / density of VS1 becomes equal. The purpose of VS2 is to confirm that the ES cells have been completely replaced with VS. This is confirmed when the oocyte / embryo has completely contracted (into the shape of a moon / kidney).

[0381] In some embodiments, human oocytes can be vitrified in groups of 2 to 4 on current freezing devices. For clinical use, standard protocols include vitrification of MII oocytes, but MI oocytes can also be vitrified. Human embryos can be vitrified at different developmental stages, including the pronuclear stage (post-fertilization), cleavage stage, and blastocyst stage. Blastocysts can be vitrified after trophectoderm biopsy for preimplantation genetic testing. Blastocysts to be tested can be vitrified individually to maintain their original identity. Blastocysts or embryos at other stages that are not to be tested can be vitrified in groups of 2 to 4.

[0382] In some embodiments, pipetting in the context of vitrification may involve the aspiration, draining, and transfer of individual or multiple oocytes and embryos, as well as bodily fluids, from one container to another. Pipetting is fundamental to IVF laboratory techniques and can be performed in a sterile manner without generating bubbles that can be destructive, lead to cell loss, or generate potentially infectious aerosols.

[0383] In some embodiments, multiple pipette types, including but not limited to Eppendorf pipettes, capillary tips, and micropipettes, can be used for different purposes during vitrification. An Eppendorf pipette is a device with a piston and spring-loaded toothed cone, a single channel, and adjustable volume (in units of 1 to 1000 μL with a specific range), used in conjunction with an Eppendorf tip to aspirate and dispense precise (usually low) volumes. A capillary tip refers to a pipette tip used in conjunction with a capillary pipette. The tips are made from flexible medical-grade plastic to prevent scratching of plastic petri dishes. Tips are manufactured in different inner diameters ranging from 75 μm to 600 μm, with the most commonly used sizes being 155 to 200 μm for oocyte detachment and processing of oocytes and embryos, and 300 μm for blastocyst processing. Micropipettes can also be used with automated robotic pipetting systems on IVF / ICSI platforms. Micropipettes, such as holding pipettes, are ultrafine glass instruments manufactured with specific outer and inner diameters for holding or manipulating oocytes and embryos during microsurgical procedures using fine-tuning devices. Micro-holding pipettes can have, for example, an outer diameter of 65–180 μm and an inner diameter of 15–30 μm.

[0384] In one exemplary embodiment of the vitrification process of an IVF / ICSI platform system, the sample hardware configuration may include the following: Calibration dish (marking positions and end positions for CM, ES, VS1, and VS2 using a diamond pen) Inverted microscope Electric stage Digital microscope (low magnification) Light sensor Stage heating and control device Fine adjustment device A range of 12,500 μm for each axis. Maximum speed 10,000μm / sec 20x objective lens 4x objective lens Stepper motor control device

[0385] In some embodiments, the IVF / ICSI platform may use a controllable electric pipette, for example, an assembly comprising: 1) functional components of an electric pipette (non-proprietary), consisting of a plunger, lubrication spring, and body / cylinder (non-proprietary); 2) a high-precision motor for driving the plunger; and 3) a single-axis electric slider (proprietary) for moving the pipette assembly in the Z direction. The components can be moved at controlled speed and acceleration using proprietary software.

[0386] In some embodiments, an example of an automated robotic vitrification sequence is described below. Each element of the sequence can be performed autonomously by the IVF / ICSI platform. 1. The system can be calibrated by assigning locations within the dish where the vitrification solution, oocytes, and embryos are placed. 2. The solutions for vitrification can be placed in the plate: equilibration solution (ES) in one depression and vitrification solution (VS) in a second depression. These depressions can serve as storage areas for the solutions for the vitrification procedure. 3. The IVF / ICSI platform can dispense 15 μL of buffer (BS) droplets into a dish and transfer vitrified oocytes / embryos into the droplets. 4. The IVF / ICSI platform allows this dish to be placed on the microscope stage along with the oocytes / embryos. 5. The automated system can initiate the vitrification protocol with a command issued locally or remotely from a computer by an operator. 6. When the stage moves to the specified coordinates, the motorized pipette with a disposable tip can be positioned above recess 1, and the pipette can descend into the recess (Z-axis) to a specified depth, and negative pressure can be applied to aspirate 30 μL of ES. When the stage moves to position the dish directly below the pipette tip, the pipette can move upward (Z-axis) to raise its tip outside and above the recess, and remain stationary. The pipette can descend onto the dish and stop just before impacting the bottom, and AI / ML can apply positive pressure to deposit the entire volume of ES contained in the tip at a predetermined position on the dish. The pipette can move upward (Z-axis) to raise its tip outside and above the recess, and remain stationary, and the stage moves to position recess 2 containing VS directly below the pipette tip. The same aspiration and discharge sequence can be repeated, for example, twice with different pre-specified volumes to generate two 100 μL droplets of VS (VS1 and VS2) in a dish. 7. The distance between the center of one droplet and the next droplet can be predetermined and programmed for specific purposes to enable the fusion of these droplets during the vitrification procedure. Multiple different distances can be used, for example, between BS and ES, between ES and VS1, and between VS1 and VS2. 8. Once the last solution is dispensed, the pipette can move upward (Z-axis), exit the droplet, return to its original position, and remain stationary for the remainder of the procedure. 9. Next, the stage moves to bring the oocyte / embryo into the field of view of the optical system of the IVF / ICSI platform. 10. Using the AI / ML described herein, an egg or embryo can be identified, and its precise location can be visualized and confirmed. 11. The micro-retaining pipette, held on the side of the stage within the tool holder, can be precisely lowered (Z-axis) near the oocyte / embryo, and the AI / ML can be automatically notified of the application of negative pressure via the motorized microinjector, allowing the oocyte / embryo to be fixed on the retaining pipette with the zona pellucida held by the negative pressure. The pressure can be automatically controlled to remain constant relative to the oocyte or embryo. 12. On the opposite side of the microtool holder, on the side of the stage, there may be a compatible tool holder for a rod-shaped instrument that serves as a holder for a freezing device ("mitogen racket"). The "racket" may be a fine mesh / grille whose surface can be used to "contain" oocytes and embryos while suspended in LN2 or N2 vapor. 13. With the oocyte immersed in BS on a holding pipette, AI / ML can be used to assist the following performance: The racket can be lowered into the same droplet and positioned on the same plane, the microtool holder can be moved forward toward the racket, slightly raised, and moved further forward until it reaches the center of the racket, and then the oocyte can be lowered onto the racket while holding the oocyte in place. From here, the holding pipette and racket holding the oocyte remain stationary together while the stage moves (X-axis), moving the entire "assembly" through the solution in a predetermined order for a predetermined time. The process of moving the oocyte or embryo / racket assembly between droplets results in "body fluid cross-linking" between droplets, which leads to droplet fusion. This makes it possible to gradually expose the oocyte / embryo to a solution with gradually increasing concentrations of cryoprotectant. 14. The oocyte or embryo / racket assembly can remain in each droplet for a predetermined time. For example, 1 minute in ES, 30 seconds in VS1, and another 30 seconds in VS2. 15. For example, after 30 seconds in VS2, AI / ML can be used to assist the movement of the stage so that the oocyte or embryo / racket assembly moves downward (Y-axis) in the dish, creating a VS2 trace and reducing the volume of VS2 on the racket and around the oocyte. 16. Next, AI / ML can be used to assist in stopping the stage in a predetermined position. 17. Next, AI / ML can be used to induce the release of the oocytes from the holding pipette onto the grid, and the racket / retainer assembly is removed and ready to be placed into LN2. 18. Next, AI / ML can be used to induce the upward movement of the holding pipette, which can then be kept stationary in place. 19. Next, AI / ML is used to induce the upward movement of the racket remaining in place, thereby allowing the racket to be removed from the fine-tuning device.

[0387] The IVF / ICSI platform can automatically remove the racket-bearing holder from the fine-tuning device and load it into the LN2 using a robotic loading system. In some embodiments of the present invention, the IVF / ICSI platform described herein may include a vitrification module that provides modifications for using an inverted microscope at a magnification that allows visualization of the sample and cells in its storage medium for operation. In some embodiments, the vitrification module may use a selection device to avoid shadows that may be generated on the magnification scale and characteristics of the cells to be selected. In some embodiments, the microscope may be automatically adjusted and moved to allow immobilization in the storage medium, as opposed to an object that carries the oocytes together with the solution. Figure 64 shows an adapter that can connect the microscope to a metal extruder (6400). Figure 65 shows an adapter for using an Eppendorf fine-tuning device with a metal extruder (6500). Figure 66 shows an adapter for using a rod-shaped light source with a metal extruder (660). Figure 67 shows a device with aspiration capability, independent Z control, and variable volume pipette, capable of visual, XYZ movement, and fine adjustment movement (6700). Figure 68 shows a device that can be attached to the neck of a microscope and has independent Z movement and pipetting capabilities (6800).

[0388] In some embodiments, the imaging and AI / ML processes of the vitrification module may provide an algorithm or a set of algorithms for determining the buoyancy capacity of cells for preservation and assigning buoyancy and depth values ​​that allow cells to be preserved.

[0389] In some embodiments, the imaging and AI / ML processes of the vitrification module can provide one or more algorithms for determining the optimal depth at which cells should be placed for preservation.

[0390] In some embodiments, the vitrification module may include a robotic mechanism for immersing a sample in liquid nitrogen and then transferring it to a cryopreservation container (e.g., a Dewar vial or tank). The robotic mechanism may include an arm, a handle, or other types of devices that can be attached to the container containing the sample and automatically move the sample into the cryopreservation device, position it at an optimal depth, or at other positional coordinates including positional coordinates derived by the AI / ML process described herein, and release the sample into the cryopreservation container. In some embodiments, the release of the sample can be achieved by separating the magnet from the magnetized sample container, and the sample is no longer magnetically coupled and is not held by the robotic device. In some embodiments, the vitrification module may include multiple robotic arms or other robotic devices for processing multiple samples simultaneously and transferring each sample to the cryopreservation device.

[0391] In some embodiments, the vitrification module can automatically pair the microscope focus on the cell sample with the tip of the pipette, so the vitrification module does not need to inform the robotic arm of its location. The robotic arm can then inform the robotic arm of the sample's location because they are paired and moving together, and therefore, if the cells move, for example, if they move deeper, this does not cause imaging or other problems for the vitrification module as they remain aligned. This process and pairing can reduce calculation errors.

[0392] In some embodiments, the cryopreservation device of the IVF / ICSI platform may include a camera that can automatically guide the robotic positioning of the specimen within the cryopreservation device and / or confirm the correct placement of the specimen.

[0393] In some embodiments of this disclosure, the vitrification and cryopreservation module may include cryopreservation components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for automated measurement and testing. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for optimization. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for system configuration and control. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the vitrification and cryopreservation module may include fully autonomous AI / ML. In some embodiments of this disclosure, the vitrification and cryopreservation module may include sensor components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include sample management components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include robotic processing systems and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include storage components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for automated measurement and testing. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for optimization. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for system configuration and control. In some embodiments of this disclosure, the vitrification and cryopreservation module may include AI / ML for semi-autonomous, supervised, or autonomous robotic mechanisms. In some embodiments of this disclosure, the vitrification and cryopreservation module may include fully autonomous AI / ML. In some embodiments of this disclosure, the vitrification and cryopreservation module may include sensor components, systems, and processes.In some embodiments of this disclosure, the vitrification and cryopreservation module may include specimen management components, systems, and processes. In some embodiments of this disclosure, the vitrification and cryopreservation module may include robotic processing systems and processes.

[0394] Referring to Figures 69 to 96, in some embodiments of this disclosure, the IVF / ICSI platform described herein may include expert systems, self-organizing systems, machine learning systems, artificial intelligence systems, etc., and may benefit from the use of neural networks such as pattern recognition, classification of one or more parameters, characteristics or phenomena, assistance for autonomous control, and neural networks trained for other purposes. References to neural networks throughout this disclosure include feedforward neural networks, radial basis function neural networks, self-organizing neural networks (e.g., Kohonen self-organizing neural networks), recurrent neural networks, modular neural networks, artificial neural networks, physical neural networks, multilayer neural networks, convolutional neural networks, hybrid neural networks with other expert systems (e.g., hybrid fuzzy logic neural network systems), autocoding neural networks, stochastic neural networks, time-delay neural networks, convolutional neural networks, modulo-feedback neural networks, radial basis function neural networks, recurrent neural networks, Hopfield neural networks, Boltzmann machine neural networks, self-organizing map (SOM) neural networks, learning vector quantization (LVQ) neural networks, fully recurrent neural networks, simply recurrent neural networks, echo-state neural networks, long short-term memory neural networks, bidirectional neural networks, hierarchical neural networks, and stochastic neural networks. Path networks, genetic scale RNN neural networks, mechanical committee neural networks, associative neural networks, physical neural networks, instantaneous learning neural networks, spiking neural networks, neocognitron neural networks, dynamic neural networks, cascaded neural networks, neuro-fuzzy neural networks, constitutive pattern generation neural networks, memory neural networks, hierarchical time memory neural networks, deep feedforward neural networks, gated recurrent unit (GCU) neural networks, autocoding neural networks, variational autocoding neural networks, denoising autocoding neural networks, sparse autocoding neural networks, Markov chain neural networks, restricted Boltzmann machine neural networks, deep belief neural networks, deep convolutional neural networks, inverse convolutional neural networks, deep convolutional inverse graphics neural networks, adversarial generative neural networks, liquid state machine neural networks, extreme learning machine neural networks, echo state neural networks, deep residual neural networks, support vector machine neural networks,It should be understood that this encompasses a wide range of different types of neural networks, machine learning systems, and artificial intelligence systems, including neural Turing machine networks and / or holographic associative memory networks, or hybrids or combinations thereof, or combinations with other expert systems such as rule-based systems and model-based systems (including those based on physical models, statistical models, flow-based models, biological models, biomimetic models, etc.).

[0395] In some embodiments, Figures 70–96 show exemplary neural networks that can be used by the IVF / ICSI platform, and Figure 69 shows a legend illustrating the various components of the neural network shown throughout Figures 70–96. Figure 69 shows various neural network components shown within cells to which function and requirements are assigned (6900). In some embodiments, examples of various neural networks may include backfeed data / sensor cells, data / sensor cells, input cells containing noise, and hidden cells. Neural network components may also include stochastic hidden cells, spiking hidden cells, output cells, input / output corresponding cells, recursive cells, memory cells, different memory cells, kernels, and convolutional or pooled cells.

[0396] In some embodiments, Figure 70 shows an exemplary perceptron neural network 7000 that can be connected to, integrated with, or interfaced with an IVF / ICSI platform. The platform can also be associated with further neural network systems such as a feedforward neural network 7100 (Figure 71), a radial basal neural network 7200 (Figure 72), a deep feedforward neural network 7300 (Figure 73), a recurrent neural network 7400 (Figure 74), a long-term / short-term neural network 7500 (Figure 75), and a gated recurrent neural network 7600 (Figure 76). The platform can also be associated with further neural network systems such as the autocoding neural network 7700 (Figure 77), variational neural network 7800 (Figure 78), denoising neural network 7900 (Figure 79), sparse neural network 8000 (Figure 80), Markov chain neural network 8100 (Figure 81), and Hopfield network neural network 8200 (Figure 82). The platform can also be further associated with additional neural network systems such as the Boltzmann machine neural network 8300 (Figure 83), restricted BM neural network 8400 (Figure 84), deep belief neural network 8500 (Figure 85), deep convolutional neural network 8600 (Figure 86), inverse convolutional neural network 8700 (Figure 87), and deep convolutional inverse graphics neural network 8800 (Figure 88). The platform can also be associated with further neural network systems such as the Adversarial Generative Network 8900 (Figure 89), the Liquid State Machine Network 9000 (Figure 90), the Extreme Learning Machine Network 9100 (Figure 91), the Echo State Network 9200 (Figure 92), the Deep Residual Network 9300 (Figure 93), the Kohonen Network 9400 (Figure 94), the Support Vector Machine Network 9500 (Figure 95), and the Neural Turing Machine Network 9600 (Figure 96).

[0397] The above neural network may have various nodes or neurons that can perform various functions in response to inputs, such as inputs received from sensors or other data sources, including other nodes. Functions may include weights, features, feature vectors, etc. Neurons may include perceptrons, neurons that mimic biological functions (such as human touch, sight, taste, hearing, and smell), etc. Continuum neurons with sigmoid activation can be used in the context of various forms of neural networks, including forms involving backpropagation.

[0398] In many embodiments, an expert system or neural network can be trained by an operator or administrator, or based on a dataset, model, etc. Training may include presenting the neural network with one or more training datasets representing values ​​such as sensor data, event data, parameter data, and other types of data (including many types described throughout this disclosure), as well as one or more indicators of results such as process results, computation results, event results, and work results. Training may include training in optimization, such as training the neural network to optimize one or more systems based on one or more optimization techniques such as Bayesian methods, parametric Bayesian classifier methods, k-nearest neighbor classifier methods, iterative methods, interpolation methods, Pareto optimization methods, and algorithmic methods. Feedback may be provided in the process of variation and selection, such as by using a genetic algorithm that evolves one or more solutions based on feedback through a series of rounds.

[0399] In some embodiments, multiple neural networks can be deployed in a cloud platform that receives data flows and other inputs collected in one or more processing environments (e.g., by mobile data acquisition devices) and transmitted to the cloud platform via one or more networks, including the use of network coding to provide efficient transmission. The cloud platform can use multiple different neural networks of various types (including modular, structurally adaptive, and hybrid forms) to perform predictive, classification, control functions and provide other outputs, as described in relation to the expert systems disclosed throughout this disclosure, potentially using massively parallel computing capabilities. Different neural networks can be structured to compete with each other so that an expert system, etc., can select the appropriate type of neural network with the appropriate input set, weights, node types, and functions, etc., for a particular task involving a given situation, workflow, environmental process, system, etc. (potentially including the use of evolutionary algorithms, genetic algorithms, etc.).

[0400] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize a feedforward neural network that moves information in one direction, such as from a data input, such as a data source related to at least one resource or parameter related to a processing environment, such as one of the data sources mentioned throughout this disclosure, through a series of neurons or nodes to an output. The data can move from the input node to the output node, possibly through one or more hidden nodes, without loops. In some embodiments, the feedforward neural network can be constructed of various types of units, such as binary McCulloch-Pitts neurons, the simplest of which is a perceptron.

[0401] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use encapsulated neural networks for predictive, classifiable, and controllable functions related to a processing environment, such as being related to one or more of the mechanical and automated systems described throughout this disclosure.

[0402] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize radial basis function (RBF) neural networks, which can be ideal in several situations involving interpolation in multidimensional space (such as when interpolation is useful for optimizing multidimensional functions, such as for optimizing the efficiency or output of data markets, power generation systems, factory systems, etc., as described herein) or other situations involving multidimensionality. In some embodiments, each neuron in the RBF neural network stores an example from a training set as an "archetype." The linearity involved in the function of this neural network generally provides the RBF with the advantage of being free from the problem of local minimums or maximums.

[0403] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can use radial basis function (RBF) neural networks, such as a distance criterion with respect to the center (e.g., a Gaussian function). Radial basis functions can be applied in place of hidden layers, such as sigmoid hidden layer transfers, in multilayer perceptrons. An RBF network may have two layers, such as a layer where the input is positioned in each RBF of the hidden layer. In some embodiments, the output layer may include a linear combination of hidden layer values, for example, representing the mean predicted output. The output layer values ​​can provide an output identical or similar to that of a regression model in statistics. In classification problems, the output layer may be a sigmoid function of a linear combination of hidden layer values, representing the posterior probability. Performance in both cases is often improved by shrinkage techniques, such as ridge regression in classical statistics. This corresponds to the prior belief in small parameter values ​​(and therefore a smooth output function) in the Bayesian framework. Since the RBF network is only tuned in the learning process by the linear mapping from the hidden layer to the output layer, local minima can be avoided. Linearity ensures that the error surface is quadratic and therefore has a single minimum. In regression problems, this can be found with a single matrix operation. In classification problems, the fixed nonlinearity introduced by the sigmoid output function can be handled using iteratively reweighted least-squares functions, etc. RBF networks can use kernel methods such as support vector machines (SVMs) and Gaussian processes (where RBF is the kernel function). Using a nonlinear kernel function, input data can be projected into a space where learning problems can be solved using linear models.

[0404] In some embodiments, an RBF neural network may include an input layer, a hidden layer, and a summation layer. In the input layer, one neuron appears for each predictor variable. For categorical variables, N-1 neurons are used, where N is the number of categories. In some embodiments, the input neurons can have their range of values ​​standardized by subtracting the median and dividing by the interquartile range. The input neurons can then supply values ​​to each of the neurons in the hidden layer. A variable number of neurons can be used in the hidden layer (determined by the training process). Each neuron may consist of a radial basis function centered at a point with the same number of dimensions as the number of predictor variables. The spread (e.g., radius) of the RBF function may differ for each dimension. The center and spread can be determined by training. Given a vector of input values ​​from the input layer, the hidden neurons can calculate the Euclidean distance from the neuron center point of the test case and then apply the RBF kernel function to this distance, for example, by using the spread value. The resulting values ​​can then be passed to the summation layer. In the summation layer, the values ​​coming from neurons in the hidden layer can be multiplied by the weights associated with the neurons and added to the weighted values ​​of other neurons. This sum becomes the output. In a classification problem, one output is generated for each target category (using a separate set of weights and summation units). The value output for a category is the probability that the example being evaluated belongs to that category. In training an RBF, various parameters can be determined, such as the number of neurons in the hidden layer, the coordinates of the center of each hidden layer function, the spread of each function in each dimension, and the weights applied to the output when it is passed to the summation layer. Clustering algorithms (such as k-means clustering) and evolutionary methods can be used for training.

[0405] In some embodiments, a recurrent neural network can have time-varying real-valued activations (outputs) (not just 0 or 1). Each connection can have modifiable real-valued weights. Some nodes are called marker nodes, some are called output nodes, and others are called hidden nodes. In supervised learning in a discrete-time setting, the training sequence of real-valued input vectors can be a sequence of activations of input nodes, one input vector at a time. At each time step, each non-input unit can compute its current activation as a nonlinear function of the weighted sum of the activations of all units that it connects to. The system can explicitly activate some output units at certain time steps (independent of the input signal).

[0406] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use self-organizing neural networks, such as Kohonen self-organizing neural networks, for visualization of data images, such as low-dimensional images of high-dimensional data. The self-organizing neural network can apply competitive learning to a set of input data, such as from or related to a processing environment, including any machine or component relating to the processing environment, such as one or more sensors or other data inputs. In some embodiments, the self-organizing neural network can be used to identify structures within data, such as unlabeled data, such as data sensed from various data sources relating to a processing environment or from sensors inside or outside the processing environment, where the source of the data is unknown (for example, when events may be coming from any of a range of unknown sources). The self-organizing neural network can organize them so that structures or patterns within data can be recognized, analyzed, and labeled, such as identifying market behavior structures as corresponding to other events and signals.

[0407] In some embodiments, methods and systems described herein, including expert systems or self-organizing functions, can use recurrent neural networks that enable bidirectional flow of data, such as when connected units (e.g., neurons or nodes) form directed cycles. Using such networks, dynamic temporal behavior can be modeled or demonstrated, such as engaging with dynamic systems, such as automated agents that interact with markets for purposes such as data collection, spot market trading testing, and execution trading, with dynamic system behavior, and complex interactions that a user may want to understand, predict, control and / or optimize, involving dynamic system behavior. For example, using a recurrent neural network, market conditions can be predicted, such as those involving dynamic processes or actions, such as changes in the state of resources traded in or enabling the market in a trading environment. In some embodiments, a recurrent neural network can use internal memory to process a set of inputs, such as inputs from other nodes and / or sensors of various types described herein, as well as other data inputs from or relating to a processing environment. In some embodiments, recurrent neural networks can also be used for pattern recognition, such as recognizing machines, components, agents, or other items based on behavioral signatures, profiles, or a set of feature vectors (e.g., audio files, images). In one non-limiting example, a recurrent neural network could recognize transitions in market or machine operating modes by learning to classify shifts from a training dataset consisting of data flows from one or more sensor data sources applied to or relating to one or more resources.

[0408] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize modular neural networks, which may include a set of independent neural networks (such as the various types described herein) coordinated by a mediator. Each independent neural network within the modular neural network may operate with separate inputs and accomplish subtasks that constitute the task that the modular network as a whole intends to perform. For example, the modular neural network may include a recurrent neural network for pattern recognition that recognizes what type of machine or system is being sensed by one or more sensors provided as input paths to the modular network, and an RBF neural network for optimizing the behavior of the machine or system once understood. The mediator may receive and process the inputs of each individual neural network and create outputs for the modular neural network, such as appropriate control parameters or state predictions.

[0409] Any combination of pairs, triplets, or larger combinations of the various neural network types described herein is encompassed by this disclosure. This may include a combination in which an expert system uses one neural network for recognizing patterns (e.g., patterns indicating a problem or failure condition) and different neural networks for self-organizing work or workflows based on the recognized patterns (e.g., providing outputs that govern the autonomous control of the system in accordance with the recognized conditions or patterns). This may also include a combination in which an expert system uses one neural network for classifying items (e.g., identifying a machine, component, or operating mode) and different neural networks for predicting the condition of an item (e.g., a failure condition, operating condition, expected condition, maintenance condition, etc.). A modular neural network may also include a situation in which an expert system uses one neural network for determining a state or situation (such as the state of a machine, process, workflow, market, memory system, network, data acquisition device, etc.) and different neural networks for self-organizing a process that includes the state or situation (e.g., a data storage process, a network coding process, a network selection process, a data market process, a power generation process, a manufacturing process, a refining process, a drilling process, a borehole process, or other processes described herein).

[0410] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use a physical neural network that performs or mimics the behavior of a nerve using one or more hardware elements. In some embodiments, one or more hardware neurons may be configured to continuously output voltage values, current values, etc., representing sensor data, such as calculating information from analog sensor inputs representing energy consumption, energy production, etc., by one or more machines, etc., that provide or consume energy for one or more processes. One or more hardware nodes may be configured to continuously output output data resulting from the activity of the neural network. Hardware nodes, which may include one or more chips, microprocessors, integrated circuits, programmable logic controllers, application-specific integrated circuits, field-programmable gate arrays, etc., may be provided to optimize machines that generate or consume energy, or to optimize another parameter of any part of any of the types of neural networks described herein. Hardware nodes may include hardware for accelerating computation (such as dedicated processors for performing basic or more advanced computations on input data and providing outputs, dedicated processors for filtering or compressing data, dedicated processors for decompressing data, dedicated processors for compressing specific files or data types (e.g., for processing image data, video streams, acoustic signals, thermal images, heatmaps, etc.)). The physical neural network can be embodied in the data acquisition device, including those that can be reconfigured by switching or routing inputs in various configurations to provide different neural network configurations within the data acquisition device to process different types of inputs (the switching and configuration may be under the control of an expert system that may include software-based neural networks located on or remotely in the data acquisition device).A physical or at least partially physical neural network may include physical hardware nodes located within a memory system, such as for storing data in machines, data storage systems, distributed ledgers, mobile devices, servers, or cloud resources, or within a processing environment, such as for accelerating input / output functions to one or more memory elements that supply data to or retrieve data from a neural network. A physical or at least partially physical neural network may include physical hardware nodes located within a network, such as for transmitting data to or from an industrial environment, such as for accelerating input / output functions to one or more network nodes within the network, or for accelerating relay functions. In some embodiments of a physical neural network, electrically adjustable resistive materials can be used to mimic the function of neural synapses. In some embodiments, the physical hardware mimics neurons, and the software mimics the neural network between neurons. In some embodiments, the neural network complements conventional algorithmic computers. They are versatile and can be trained to perform appropriate functions without requiring any instruction, such as classification, optimization, pattern recognition, control, selection, and evolution.

[0411] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can use multilayer feedforward neural networks for the classification of complex patterns such as one or more items, phenomena, modes, or states. In some embodiments, the multilayer feedforward neural network can be trained by optimization techniques such as genetic algorithms to explore a large and complex space of choices to find an optimal or near-optimal overall solution. For example, one or more genetic algorithms can be used to train a multilayer feedforward neural network to classify complex phenomena, such as recognizing complex operating modes of machines, including modes with complex interactions between machines (including interference effects, resonance effects, etc.), modes with nonlinear phenomena, and modes with critical failures such as multiple simultaneous failures that make root cause analysis difficult. In some embodiments, the multilayer feedforward neural network can be used to classify results from market monitoring, such as monitoring systems such as automated agents operating within the market, and from monitoring of resources that enable the market, such as computing, networking, energy, data storage, energy storage, and other resources.

[0412] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can use multilayer perceptron (MLP) neural networks with feedforward and backpropagation methods to process one or more remote sensing applications or to acquire input from sensors distributed across various processing environments. In some embodiments, the MLP neural network can be used to classify market and resource environments such as spot markets, futures markets, energy markets, renewable energy credit (REC) markets, networking markets, advertising markets, spectral markets, ticketing markets, reward markets, computing markets, and other markets mentioned throughout this disclosure, as well as physical resources and environments that generate them, such as energy resources (including renewable energy environments, mining environments, exploration environments, drilling environments, etc.), including the classification of geological structures (including subsurface and surface features), the classification of materials (including fluids, minerals, metals, etc.), and other issues, as well as the classification of physical resources and environments that generate them, such as energy resources (including renewable energy environments, mining environments, exploration environments, drilling environments, etc.). This may include fuzzy classification.

[0413] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize structure-adaptive neural networks, the structure of which is adapted based on rules, perceived states, situational parameters, etc. For example, if a neural network does not converge to a solution, such as classifying items or reaching a prediction, when acting on a set of inputs after a certain amount of training, the neural network can be changed from a feedforward neural network to a recurrent neural network, for example, by switching data paths between some subsets of nodes from unidirectional to bidirectional data paths. Structural adaptation can be performed under the control of an expert system to induce adaptation when triggers, rules, or events occur, such as recognizing the occurrence of a threshold (e.g., no convergence to a solution within a given time) or recognizing a phenomenon as requiring a different or additional structure (e.g., recognizing that the system is changing dynamically or nonlinearly). In a non-limiting example, an expert system, upon receiving indication that continuously variable transmission is being used to drive generators, turbines, etc., within the system being analyzed, may switch from simpler neural network structures, such as feedforward neural networks, to more complex neural network structures, such as recurrent neural networks or convolutional neural networks.

[0414] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use autoencoders, autoassociators, or diabolo neural networks that may be analogous to multilayer perceptron (MLP) neural networks, such that there may be input layers, output layers, and one or more hidden layers connecting them. However, the output layer in an autoencoder may have the same number of units as the input layer, and the purpose of an MLP neural network is to reconstruct its own input (rather than simply emitting a target value). Thus, an autoencoder can operate as an unsupervised learning model. An autoencoder can be used for unsupervised learning of efficient coding, such as dimensionality reduction or learning a data generation model. In some embodiments, an autoencoder neural network can be used to self-learn efficient network coding for the transmission of analog sensor data from a machine or digital data from one or more data sources over one or more networks. In some embodiments, an autoencoder neural network can be used to self-learn efficient memory techniques for storing data flows.

[0415] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize a stochastic neural network (PNN), which in some embodiments may include a multilayer (e.g., four-layer) feedforward neural network, where the layers may include an input layer, a hidden layer, a pattern / summation layer, and an output layer. In one embodiment of the PNN algorithm, the parent probability distribution function (PDF) of each class can be approximated by a Parzen window and / or a nonparametric function, etc. Then, using the PDF of each class, Bayes' rules can be employed to estimate the class probability of a new input and assign it to the class with the highest posterior probability. The PNN can embody a Bayesian network and may utilize statistical algorithms or analytical techniques such as kernel Fisher discriminant analysis techniques. The PNN can be used for classification and pattern recognition in any of the extensive embodiments disclosed herein. In a non-limiting example, a stochastic neural network can be used to predict engine failure conditions based on the collection of data inputs from engine sensors and equipment.

[0416] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize a time-delayed neural network (TDNN), which may include a feedforward structure for sequence data that recognizes features independently of sequence position. In some embodiments, delays are added to one or more inputs or between one or more nodes so that multiple data points (from different time points) are analyzed together to account for time shifts in the data. The time-delayed neural network may form part of a larger pattern recognition system, such as using a perceptron network. In some embodiments, the TDNN may be trained in supervised learning, such as when the coupled weights are trained under backpropagation or feedback. In some embodiments, the TDNN may be used to process sensor data from different flows, such as a flow of velocity data, a flow of acceleration data, a flow of temperature data, or a flow of pressure data, and the time delay may be used to temporally align the data flows, such as helping to understand patterns with an understanding of various flows (e.g., changes in price patterns in a spot or futures market).

[0417] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize convolutional neural networks (sometimes referred to as CNNs, ConvNets, shift-invariant neural networks, or spatially invariant neural networks), where units are connected in a pattern similar to the visual cortex of the human brain. Neurons can respond to stimuli in a limited area of ​​space called a receptive field. Receptive fields may partially overlap so that they collectively cover an entire (e.g., visual) area. Node responses can be mathematically computed, such as by convolutional operations using multilayer perceptrons with minimal preprocessing. Convolutional neural networks can be used for recognition in image and video streams, such as using camera systems placed on mobile data acquisition devices such as drones and mobile robots to recognize types of machines in a large environment. In some embodiments, convolutional neural networks can be used to provide recommendations based on data inputs, including sensor inputs and other contextual information, such as recommending routes for mobile data acquisition devices. In some embodiments, convolutional neural networks can be used to process inputs such as natural language processing of instructions provided by one or more parties involved in a workflow within an environment. In some embodiments, the convolutional neural network can be deployed using a large number of neurons (e.g., 100,000, 500,000 or more), multiple layers (e.g., 4, 5, 6 or more), and a large number of parameters (e.g., millions). The convolutional neural network can use one or more convolutional networks.

[0418] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use a coordinating feedback network to recognize emergency phenomena (such as new types of behavior not previously understood in the processing environment).

[0419] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize self-organizing maps (SOMs) that include unsupervised learning. A set of neurons can learn to position points in an input space onto coordinates in an output space. The input space may have different dimensions and forms than the output space, and the SOM can store these while positioning phenomena into groups.

[0420] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize a learning vector quantized neural network (LVQ). Representatives of class archetypes can be parameterized in a distance-based classification system along with an appropriate distance measure.

[0421] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize echo-state networks (ESNs) that may include recurrent neural networks having loosely connected, disordered hidden layers. The weights of the output neurons can be modified (for example, the weights can be trained based on feedback). In some embodiments, ESNs can be used to handle time-series patterns, such as recognizing patterns of market-related events, for example, patterns of price changes in response to stimuli.

[0422] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use a bidirectional recurrent neural network (BRNN) to predict or label each element of an array based on both the past and future context of the elements, such as using a finite array of values ​​(e.g., voltage values ​​from a sensor). This can be done by adding the outputs of two RNNs, such as one processing the array from left to right and the other from right to left. The combined output is a prediction of a target signal, such as one provided by a teacher or supervisor. The bidirectional RNN can be combined with a long-term memory RNN.

[0423] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use a hierarchical RNN that connects elements in various ways to decompose hierarchical behavior into useful subprograms or the like. In some embodiments, a hierarchical RNN can be used to manage one or more hierarchical templates for data collection in a processing environment.

[0424] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can utilize stochastic neural networks, thereby introducing disordered fluctuations into the network. Such disordered fluctuations can be considered a form of statistical sampling, such as Monte Carlo sampling.

[0425] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can utilize genetically scaled recurrent neural networks. In some of these embodiments, RNNs (often LSTMs) are used, and the sequence is decomposed into several scales, all of which represent the main length between two consecutive points. A first-order scale consists of a regular RNN, a second-order scale consists of all points separated by two exponents, and so on. An Nth-order RNN connects the first and last nodes. The outputs from all the different scales can be processed as a committee of members, and the associated scores can be used genetically in subsequent iterations.

[0426] In some embodiments, methods and systems described herein that include expert systems or self-organizing functions may use a Committee of Machines (CoM) comprising a set of different neural networks that “vote” together on a given example. Since neural networks can suffer from local minima, starting with the same structure and training but using different initial weights randomly often yields different results. CoMs tend to stabilize the results.

[0427] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can utilize associative neural networks (ASNNs), such as those including a mechanical committee-type extension combining multiple feedforward neural networks and k-neighborhood techniques. It can use correlations between ensemble responses as a measure of distance between analyzed cases, compared to kNNs. This corrects for bias in the neural network ensemble. The associative neural network can have memory that can match the training set. As new data becomes available, the network instantly improves its predictive ability, providing data approximation (self-learning) without retraining. Another important feature of ASNNs is the possibility of interpreting the neural network's results by analyzing correlations between data cases in the model space.

[0428] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may use an instantaneous learning neural network (ITNN) in which the weights of the hidden and output layers are directly located from the training vector data.

[0429] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, may utilize spiking neural networks that can explicitly consider the timing of inputs. Network inputs and outputs can be represented as a series of spikes (e.g., delta functions, more complex shapes). SNNs can process time-domain information (e.g., signals that change over time, such as signals with dynamic behavior in a market or processing environment). They are often implemented as recurrent networks.

[0430] In some embodiments, the methods and systems described herein, including expert systems or self-organizing functions, can use dynamic neural networks that address nonlinear multivariate behavior and include learning of time-dependent behaviors such as transients and delay effects. Transient states may include behaviors that change market variables such as price, available quantity, and available counterparti...

Claims

1. A method for robotic pipetting based on artificial intelligence for cell preparation, Training an artificial intelligence / machine learning system (AI / ML system) to classify images of biological specimens, wherein the classification is partially based on cellular material in the image that conforms to a target cell criterion. To store the classified images, Receiving an image object from an imaging system comprising a microscope system, a camera system configured to receive imaging from the microscope system, and an illumination system configured to illuminate biological material, wherein the microscope system, camera system and illumination system are operably connected, and the image object satisfies a predetermined threshold based in part on the classified image. Processing the image object using the AI / ML system that determines the presence of recoverable target cells. Position the robotic pipette in the target physical orientation relative to the target cells that can be recovered. The location of the robotic pipette is confirmed using an AI / ML system in the physical direction of the target, and To instruct the robotic pipette to retrieve the recoverable target cells from the physical direction of the target. Methods that include...

2. The method according to claim 1, wherein the image object is a still image.

3. The method according to claim 1, wherein the image object is an image taken over time.

4. The method according to claim 1, wherein the image object is a video.

5. The method according to claim 1, wherein the predetermined threshold is the probability that at least a portion of the image objects includes the target classified image type.

6. The method according to claim 5, wherein the classified image type to be targeted is an image of a classified cell type.

7. The method according to claim 1, wherein the use of an AI / ML system for determining the presence of recoverable target cells includes repeatedly acquiring a plurality of image objects until at least one of the plurality of image objects satisfies a predetermined threshold.

8. The method according to claim 7, wherein repeatedly acquiring the plurality of image objects includes changing the physical arrangement of the imaging system to acquire a plurality of image orientations from the plurality of image objects.

9. The method according to claim 7, wherein repeatedly acquiring the plurality of image objects includes changing the type of illumination used by the imaging system to acquire a plurality of image orientations within the plurality of image objects.

10. The method according to claim 9, wherein the illumination type is polarized.

11. The method according to claim 1, wherein the recoverable target cell is a single cell.

12. The method according to claim 1, wherein the AI / ML system uses at least one neural network.

13. The method according to claim 1, wherein each component of the imaging system component includes at least one motor for positioning.

14. The method according to claim 1, wherein the physical orientation of the target is obtained in part on the basis of rotating the stage in which the recoverable target cells are located.

15. A method for robotic pipetting based on artificial intelligence for sperm preparation, To place a container containing a semen sample, Receiving image objects from an imaging system including a microscope system, camera system, and lighting system. The semen liquefaction state of the semen sample is determined based at least partially on the image object by using an artificial intelligence / machine learning system (AI / ML system) to verify the validity of the determined semen liquefaction state for sperm collection by comparing the determined semen liquefaction state with a stored liquefaction state criterion. To determine the sperm characteristics in the semen sample based at least partially on the aforementioned image object, Using an AI / ML system, the determined sperm characteristics and motility are compared with stored sperm standards to confirm the validity of the determined sperm characteristics and motility for sperm collection. Determining the speed of sperm motility based at least partially on the aforementioned image object, Using an AI / ML system, the determined sperm motility rate is compared with a stored motility standard to confirm the validity of the determined sperm motility rate for sperm separation. Once the determined liquefaction state, determined sperm characteristics and motility, and the validity of the determined sperm motility are confirmed, the robotic pipette is instructed to collect sperm from the semen sample, and To collect at least one sperm into the robotic pipette. Methods that include...

16. The method according to claim 15, wherein the arrangement of the vessel is at least partially performed by rotating the vessel using a staging system controlled by the AI / ML system.

17. The method according to claim 15, wherein the velocity of the sperm motility is at least one of the vertical motion measurement, the horizontal motion measurement, or the angular motion measurement.

18. The method according to claim 15, wherein the sperm characteristic is at least one of sperm count, motility parameters, or morphological measurements.

19. The method according to claim 15, wherein the AI / ML system uses at least one neural network.

20. The method according to claim 15, wherein the sperm collected by the robotic pipette is a plurality of sperm.

21. A method for robotic pipetting based on artificial intelligence for sperm preparation, The method involves placing a container containing a semen sample on a staging mechanism, wherein the staging mechanism includes at least one motor for positioning the container in a specified direction using an artificial intelligence / machine learning system (AI / ML system). The position of the instrument is confirmed using the AI / ML system. The AI / ML system instructs the robotic pipette to position the pipette tip within the semen sample at a location specified by the AI / ML system. The semen sample is aspirated and mixed using the robotic pipette. Receiving image objects from an imaging system including a microscope system, camera system, and lighting system. Using the AI / ML system, the liquefaction state of the semen sample is determined at least partially based on the image object. Using the AI / ML system, the determined semen liquefaction state is compared with a predetermined liquefaction threshold to confirm the validity of the determined semen liquefaction state for sperm collection. Using the robotic pipette, the semen sample is mixed according to the evaluation of the liquefaction state, and Using the robotic pipette, obtain at least one sperm from the semen sample at a location specified by the AI / ML system. A method that includes this.

22. The method according to claim 21, wherein the specified direction is the degree of inclination.

23. The method according to claim 21, wherein the specified direction is the degree of rotation.

24. The method according to claim 21, wherein the specified direction is the velocity of motion.

25. The method according to claim 21, wherein the specified direction is algorithmically determined by the AI / ML system.

26. The method according to claim 21, wherein the image object is a still image.

27. The method according to claim 21, wherein the image object is an image taken over time.

28. The method according to claim 21, wherein the image object is a video.

29. The method according to claim 21, wherein the aspiration and mixing of the semen is repeatedly performed using the AI / ML system and the imaging system to generate the plurality of image objects until the predetermined liquefaction threshold is satisfied by at least one image object of the plurality of image objects.

30. The method according to claim 21, wherein the mixing includes the addition of a culture medium containing a liquefaction reducing agent.

31. The method according to claim 21, wherein the mixture does not include the addition of a culture medium containing a liquefaction reducing agent.

32. The method according to claim 21, wherein the semen sample is placed in the layered culture medium inside the container.

33. A method for robotic pipetting based on artificial intelligence for sperm preparation, The first container containing the semen sample is placed inside the staging mechanism. Receiving an image object representing sperm in the semen sample from an imaging system including a microscope system, a camera system, and a lighting system. Using an artificial intelligence system / machine learning system (AI / ML system), the plurality of sperm within the image object are graded according to the degree to which each sperm corresponds to a sperm quality standard. Selecting high-grade sperm from the aforementioned graded sperm using the AI / ML system, The aforementioned high-grade sperm are collected in a second container using the robotic pipette and isolated. Using the AI / ML system, move the staging mechanism and position the second container at a designated location for inspection by the imaging system. Using the AI / ML system, instruct the immobilization device to block the tail function of the high-grade sperm, wherein the blockage of the tail function reduces the motility of the high-grade sperm. The robotic pipette is guided to collect the high-grade sperm by positioning the pipette tip in the second container at a position specified by the AI / ML system, and Using the robotic pipette, the high-grade sperm is collected from the second container at a location specified by the AI / ML system. A method that includes this.

34. The method according to claim 33, wherein the first image object is at least one of a single image, a time-series image, or a video.

35. The method according to claim 33, wherein the decrease in the mobility of the high-grade sperm is automatically measured by the AI / ML system.

36. The method according to claim 35, wherein the AI / ML system is used to determine sufficient immobilization by comparing the automatic measurement of the decrease in mobility with a predetermined threshold.

37. The method according to claim 33, wherein the high-quality sperm is a plurality of sperm.

38. The method according to claim 33, wherein the immobilization device is a laser.

39. The method according to claim 33, wherein the immobilization device is a pipette.

40. The method according to claim 33, wherein the immobilization device is a piezoelectric actuation device.

41. A method for robotic pipetting based on artificial intelligence for sperm preparation, To place a container containing a semen sample inside the staging mechanism, Using a robotic pipette, create at least two droplets in the dish on the staging mechanism. Using the robotic pipette, connect the at least two droplets to the culture medium pathway according to a programmable design commanded by an artificial intelligence / machine learning system (AI / ML system). Using the robotic pipette, a certain amount of sperm is deposited from the semen sample into one of the at least two droplets. The AI / ML system is used to optically scan a certain amount of sperm and generate a first image object and a second image object using an imaging system including a microscope system, a camera system and an illumination system, wherein the creation of the first image object and the second image object is separated by a specified period of time. Using the AI / ML system, compare at least one sperm component of a certain amount of sperm in the first image object with at least one sperm component of a certain amount of sperm in the second image object to calculate the real-time motility index of multiple sperm in the certain amount of sperm, at least partially based on the change in the location of the sperm between the first image object and the second image object, and To rate multiple sperm based at least partially on the real-time motility index. A method that includes this.

42. The method according to claim 41, wherein the first image object and the second image object are still images.

43. The method according to claim 41, wherein the first image object and the second image object are images taken over time.

44. The method according to claim 41, wherein the first image object and the second image object are video.

45. The method according to claim 41, wherein a certain amount of sperm is obtained without seminal plasma components.

46. The method according to claim 41, wherein one of the at least two droplets is composed of untreated seminal plasma that does not contain culture medium.

47. The method according to claim 41, wherein the semen sample is diluted by mixing the semen sample with a culture medium.

48. The method according to claim 41, wherein an optimal sperm preparation concentration is maintained in the at least two droplets by blocking the culture medium pathway, at least partially based on automatically blocking the culture medium pathway using the tip of a pipette.

49. The method according to claim 41, wherein the imaging system uses autofocus.

50. The method according to claim 46, wherein the plurality of robotic pipettes are arranged at multiple angles with respect to the position of the instrument specified by the AI / ML system.

51. The method according to claim 41, wherein the at least one sperm component is a tail.

52. The method according to claim 41, wherein the at least one sperm component is an acrosome.

53. The method according to claim 41, wherein the at least one sperm component is the head.

54. The method according to claim 41, wherein the at least one sperm component is the middle piece portion.

55. The method according to claim 41, wherein the comparison is performed at least in part on an algorithmic comparison of at least one pixel change in the second image object with respect to the first image object.

56. The method according to claim 41, wherein the comparison is performed continuously.

57. A method for robotic pipetting based on artificial intelligence for sperm preparation, An artificial intelligence / machine learning system (AI / ML system) is used to optically scan a semen sample and generate a first image object and a second image object using an imaging system including a microscope system, a camera system and an illumination system, wherein the creation of the first image object and the second image object occupies the same focal space and their creation is separated by a specified period. Comparing the first image object and the second image object using the AI / ML system, wherein the AI / ML system algorithmically measures at least one pixel change between the first image object and the second image object, Using the algorithmic measurement described above, estimate a plurality of real-time sperm motility indices for at least one subset of sperm present in the first image object and the second image object. Identifying at least one sperm having a real-time sperm motility index that satisfies a predetermined threshold, Using the AI / ML system, control the imaging system to automatically focus on the at least one sperm. Positioning the robotic pipette in the target physical orientation relative to at least one sperm, and Using the robotic pipette, obtain the at least one sperm from the semen sample at a location specified by the AI / ML system. A method that includes this.

58. The method according to claim 57, wherein the semen sample is untreated seminal plasma that does not contain a culture medium.

59. The method according to claim 57, wherein the first image object and the second image object are images taken over time.

60. The method according to claim 57, wherein the first image object and the second image object are video.

61. A robotic vitrification method based on artificial intelligence, Place at least one oocyte in a buffer or CPA solution. Using an AI / ML system, process image objects generated by an imaging system to determine the location of retrievable target oocytes. Position the robotic pipette in the target physical orientation relative to the target oocytes that can be recovered. Using the AI / ML system, the location of the robotic pipette is confirmed in the physical direction of the target. Initiating contact with the recoverable target oocytes and instructing the robotic pipette to fix the oocytes to the robotic pipette by applying negative pressure, Using a robotic microtool holder, the vitrification assembly or any other type of freezing device is lowered to a position close to the physical orientation of the target within the buffer or CPA solution dish. Using the robotic pipette, place the oocytes onto the vitrification assembly or other freezing device in the dish containing droplets of the buffer or CPA solution. The relative physical arrangement of the oocytes and the vitrification assembly / freezing apparatus remains unchanged, and the AI / ML system is used to hold the robotic pipette and the robotic microtool holder stationary relative to each other so as to form a vitrification assembly, and The AI / ML system moves the vitrified assembly by robot through different solutions for a predetermined time in a predetermined order, and generates fluid crosslinks between two or more bound droplets containing a buffer or CPA until the vitrified assembly reaches a predetermined threshold concentration of the cryoprotectant. A method that includes this.

62. The method according to claim 61, wherein the at least one oocyte is an embryo.

63. The method according to claim 61, wherein the oocyte is a plurality of embryos.

64. The method according to claim 61, wherein the image object is a still image.

65. The method according to claim 61, wherein the image object is an image taken over time.

66. The method according to claim 61, wherein the image object is a video image.

67. The method according to claim 61, wherein negative pressure is applied via an electric microinjector that is at least partially controlled by the AI / ML system.

68. The method according to claim 61, wherein the robotic pipette is a plurality of robotic pipettes.

69. The method according to claim 61, wherein the physical orientation of the target is obtained in part on the basis of rotating the stage in which the recoverable target oocytes are located.

70. The method according to claim 61, wherein the vitrified assembly is a mitogen racket.

71. The method according to claim 61, wherein the AI / ML system uses at least one neural network.

72. A robotic vitrification method based on artificial intelligence, Using an artificial intelligence system / machine learning system (AI / ML system), automatically commands are issued to a robotic pipette to obtain oocytes from a container. The robotic pipette containing the oocytes is guided into a droplet containing a freezing device or vitrification assembly held by a robotic microtool at least partially controlled by the AI / ML system. The robotic pipette is positioned in close proximity to the vitrification assembly to create a vitrification assembly-pipette combination, wherein the positioning is at least partially based on the analysis of at least one image object created by an imaging system using the AI / ML system, and the imaging system includes a microscope system, a camera system, and an illumination system. The combination of the vitrification assembly and pipette allows a robot to change the concentration of the cryoprotectant in the solution holding the oocytes, and Exposing the oocytes, held by the vitrification assembly-pipette combination, to gradually increasing concentrations of cryoprotective agent until the oocytes reach a predetermined threshold concentration of the cryoprotective agent. A method that includes this.

73. The method according to claim 72, wherein the oocyte is an embryo.

74. The method according to claim 72, wherein the oocyte is a plurality of embryos.

75. The method according to claim 72, wherein the concentration of the cryoprotectant is automatically evaluated using the AI / ML system and the imaging system.

76. The method according to claim 72, wherein the vitrified assembly is a mitogen racket.

77. The method according to claim 72, wherein the exposure of the oocytes to a gradually increasing concentration of cryoprotective agent is automatically performed according to a predetermined exposure rate by robotic control of the AI / ML system and the vitrification assembly-pipette combination.

78. The method according to claim 72, wherein the exposure rate of the oocytes to the gradually increasing concentration of the cryoprotective agent is automatically determined in real time by the AI / ML system, at least in part, based on the output of the imaging system.

79. The method according to claim 72, wherein the image object is a still image.

80. The method according to claim 72, wherein the image object is a video.

81. A method for COC recovery based on automated artificial intelligence, Using an artificial intelligence / machine learning system (AI / ML system), optically scan a follicular fluid sample and generate an image object using an imaging system including a microscope system, camera system, and illumination system. The image object is compared with a predetermined threshold using an AI / ML system, wherein the predetermined threshold is at least partially an optical pattern having a probability corresponding to a cumulus-oocyte complex (COC). Identifying the COC in the follicular fluid sample based at least partially on the image object that satisfies the predetermined threshold, Determining the location of COC in the follicular fluid sample, at least in part, based on identifying a region of the image object corresponding to an optical pattern within the image object that satisfies the predetermined threshold, Instructing the robotic pipette to collect the COC at the location of the COC, and The COC is collected in the robotic pipette. Methods that include...

82. The method according to claim 81, wherein optical scanning of the follicular fluid sample is performed automatically using at least partially the AI / ML system and the imaging system.

83. The method according to claim 81, wherein optical scanning of the follicular fluid sample is performed automatically according to a predetermined scanning pattern at least partially specified by the AI / ML system.

84. The method according to claim 81, wherein the illumination system uses multiple wavelengths of light to image the follicular fluid sample.

85. The method according to claim 81, wherein the image object is a still image.

86. The method according to claim 81, wherein the image object is a time-series image.

87. The method according to claim 81, wherein the image object is a video.

88. The method according to claim 81, wherein the region of the image object corresponding to the optical pattern is a plurality of regions.

89. The method according to claim 81, wherein the robotic pipette is a plurality of robotic pipettes.

90. The method according to claim 81, wherein the COC is a plurality of COCs.

91. A method for COC recovery based on automated artificial intelligence, Place the container containing the follicular fluid sample on the motorized stage of the microscope. The follicular fluid sample is scanned by a robot using an imaging system and an artificial intelligence / machine learning system (AI / ML system) to create an image object, wherein the imaging system includes a microscope system, a camera system and a lighting system. Identifying cumulus-oocyte complexes (COCs) and their locations within an image object, at least partially based on comparing the image object with a predetermined threshold using an AI / ML system, wherein the predetermined threshold is an optical pattern having a probability corresponding to a COC cluster. Instructing the robotic pipette to acquire the COC in the robotic pipette at the location, Using the robotic pipette, aspiration is initiated, at least partially based on applying negative pressure to the COC and adjacent bodily fluids, wherein the amount of bodily fluid aspirated is at least partially determined by the physical limits outside the COC detected within the image object. To remove the robotic pipette from the apparatus, Move the motorized stage to bring the washing dish containing the processing medium close to the robotic pipette. Lowering the robotic pipette into the washing area of ​​the washing dish, and The method involves applying positive pressure to the robotic pipette to discharge the COC and adjacent body fluids into the culture medium, wherein the discharged volume is controlled at least partially using the AI / ML system, and as a result, the positive pressure stops when the COC mass exits the robotic pipette. A method that includes this.

92. The method according to claim 91, wherein the container and the washing dish are a single common container.

93. The method according to claim 91, wherein the microscope is an inverted microscope.

94. The method according to claim 91, wherein the microscope is a movable microscope.

95. The method according to claim 91, wherein the amount of bodily fluid aspirated is determined at least partially by the AI / ML system.

96. The method according to claim 91, wherein the identified COC is a plurality of COCs.

97. The method according to claim 91, wherein the robotic pipette is a plurality of robotic pipettes.

98. The method according to claim 91, wherein scanning the follicular fluid sample is performed automatically using at least partially the AI / ML system and the imaging system.

99. The method according to claim 91, wherein the suction continues until the AI / ML system and the imaging system confirm completion.

100. The method according to claim 99, wherein the confirmation of completion is performed using the AI / ML and imaging system to confirm, at least partially, that the COC mass is sufficiently separated from other substances, based on a second image taken during or after aspiration.

101. A method for peeling based on automated artificial intelligence, The method involves scanning a dish containing a cumulus-oocyte complex (COC) with a robot using an imaging system and an artificial intelligence / machine learning system (AI / ML system) to create an image object, wherein the imaging system includes a microscope system, a camera system and a lighting system, and the dish contains an enzyme that removes cumulus cells from oocytes in the COC. Identifying the COCs and their locations within the image object, at least in part, based on comparing the image object with a predetermined threshold using an AI / ML system, wherein the predetermined threshold is an optical pattern having a probability corresponding to a COC cluster. Instructing the robotic pipette to collect the COC at the location of the COC, and The AI / ML system uses a second image object generated by the imaging system to collect the COC into the robotic pipette, discharge the COC from the pipette, and return it to the dish, repeating this process until sufficient removal of the cumulus oophorus and corona radiata cells from the COC is confirmed. Methods that include...

102. The method according to claim 101, wherein the plate includes a plurality of indentations.

103. The method according to claim 101, wherein the dish is a tube.

104. The method according to claim 101, wherein the microscope system includes a stereomicroscope.

105. The method according to claim 101, wherein the microscope system includes an inverted microscope.

106. The method according to claim 101, wherein the microscope system includes a movable microscope.

107. The method according to claim 101, wherein the enzyme is a plurality of enzymes.

108. The method according to claim 101, wherein the enzyme is a plurality of compounds.

109. The method according to claim 101, wherein scanning is performed automatically based on a predetermined pattern, at least partially using the AI / ML system and the imaging system.

110. The method according to claim 101, wherein the scanning is automatically adjusted to use a scanning pattern based at least in part on the AI / ML system and the imaging system, and the AI / ML system evaluation of at least one imaging system output.

111. The method according to claim 101, wherein the robotic pipette is a plurality of robotic pipettes.

112. The method according to claim 101, wherein repeatedly collecting the COC into the robotic pipette is performed to continuously reduce the inner diameter of the pipette used repeatedly by the robotic pipette.

113. The method according to claim 101, wherein the COC mass is a plurality of COC masses.

114. A method for preparing human eggs based on automated artificial intelligence, The method involves scanning a dish containing a cumulus-oocyte complex (COC) with a robot using an imaging system and an artificial intelligence / machine learning system (AI / ML system) to create a first image object, wherein the imaging system includes a microscope system, a camera system and a lighting system, and the dish contains an enzyme that removes cumulus cells from oocytes in the COC. Identifying the COCs and their locations within the image object, at least in part, based on comparing the first image object with a predetermined threshold using an AI / ML system, wherein the predetermined threshold is at least an optical pattern having a probability corresponding to a COC cluster. The first robotic pipette is instructed to collect the COC at the location of the COC. Using the AI / ML system, determine the first aspirated volume of oocytes / body fluid into the first robotic pipette. Using the AI / ML system and imaging system, determine the presence of air bubbles in the first aspirated volume of oocytes / body fluid. The first aspirated volume of oocytes / body fluid is discharged into a receptor placed on the stage. The AI / ML system and imaging system are used to guide the positioning of a second robotic pipette to aspirate a subset of the discharged first aspirated volume of oocytes / body fluid, wherein the second robotic pipette has a narrower inner diameter than the first robotic pipette, and Using multiple gradually narrowing robotic pipettes, the aforementioned amount of oocyte / body fluid is aspirated and discharged until sufficient detachment is confirmed using a second image object by the AI / ML system and imaging system for at least one oocyte in the aforementioned amount of oocyte / body fluid. A method that includes this.

115. The method according to claim 114, wherein the second robotic pipette has the same inner diameter as the first robotic pipette.

116. The method according to claim 114, wherein the stage is an inverted microscope stage.

117. The method according to claim 114, wherein the scanning is performed automatically according to a predetermined scanning pattern specified at least partially by the AI / ML system.

118. The method according to claim 114, wherein the first image object and the second image object are still images.

119. The method according to claim 114, wherein the first image object and the second image object are time-series images.

120. The method according to claim 114, wherein the first image object and the second image object are video.

121. A method for ICSI based on automated artificial intelligence, To receive at least one droplet containing an egg cell into a dish placed on the stage, The method involves detecting the zona pellucida using an artificial intelligence / machine learning system (AI / ML system) and an imaging system, wherein the imaging system includes a microscope system, a camera system, and an illumination system. The aforementioned egg is held using a robotic microtool. A robotic pipette is lowered into the droplet. Using the AI / ML system and imaging system, determine the region holding the egg, and position the robotic microtool in that region. Using the AI / ML system and imaging system, instruct the robotic microtool to apply negative pressure to hold the oocyte in the robotic pipette. Using the aforementioned AI / ML system and imaging system, determine the target location where zona pellucida excision should be performed. To move the egg to the target location, Using the AI / ML system and imaging system, the thickness of the zona pellucida is evaluated, and the excision operation is determined. To generate a laser to excise a predetermined area and depth of the zona pellucida, and Using the AI / ML system and imaging system, define the injection route into the egg. Methods that include...

122. The method according to claim 121, wherein the stage is a microscope stage.

123. The method according to claim 122, wherein the microscope stage is in close proximity to the inverted microscope.

124. The method according to claim 122, wherein the microscope stage is in close proximity to the stereomicroscope.

125. The method according to claim 122, wherein the microscope stage is in close proximity to the movable microscope.

126. The method according to claim 121, wherein the microscope stage is in close proximity to the optical coherence tomography apparatus.

127. The method according to claim 121, wherein the microscope stage is in close proximity to the optical interference microscope apparatus.

128. The method according to claim 121, wherein the microscope stage is in close proximity to the lensless microscope.

129. The method according to claim 121, wherein the robotic microtool is a robotic holding pipette.

130. The method according to claim 121, wherein the robotic pipette is a plurality of pipettes.

131. The method according to claim 121, wherein the plate is heated.

132. The method according to claim 121, wherein the imaging system can generate at least one image having a composite reality in which a simulated image and a real image are combined.

133. The method according to claim 121, wherein the excision operation is performed with the necessary strength and radius to excise an appropriate portion of the zona pellucida without deforming the egg, in order to facilitate the entry of the ICSI needle through the zona pellucida.

134. A method for ICSI based on automated artificial intelligence, Receiving eggs in a dish placed on the microscope stage, Using an artificial intelligence / machine learning system (AI / ML system) and an imaging system, a combination of positive and negative pressure is applied to a robotic ICSI needle to position sperm at a desired location. The robotic ICSI needle is advanced into the egg cell at a predetermined controlled speed, and stops when it reaches the end point of the path specified by the AI / ML system. The method involves using piezoelectric pulses to destroy the egg membrane (egg cell membrane), and the AI / ML system and the imaging system can be used to determine whether piezoelectric pulses are necessary to destroy the membrane and how many pulses are needed. Using the AI / ML system and the imaging system, positive pressure is applied to the robotic ICSI needle to deposit the sperm into the egg. Using the AI / ML system and the imaging system, confirm that the sperm are outside the needle, move the needle outside the egg, and Using the AI / ML system and the imaging system, positive pressure is applied to the holding pipette until the egg is released into the droplet. Methods that include...

135. The method according to claim 134, wherein the destruction of the egg membrane is performed mechanically.

136. The method according to claim 134, wherein the positive pressure within the ICSI needle pushes the sperm toward the tip of the ICSI needle in preparation for injection into the cytoplasm, thereby reducing the amount of foreign culture medium injected into the egg.

137. The method according to claim 134, wherein the desired position is determined in part on the basis of using conventional microscopy.

138. The method according to claim 134, wherein the desired position is determined in part on the basis of using optical coherence tomography.

139. The method according to claim 134, wherein the desired position is determined in part on the basis of using optical interference microscopy.

140. The method according to claim 134, wherein the desired position is determined in part on the basis of using a three-dimensional simulation of the morphology of the egg.

141. A method for identifying oocytes based on automated artificial intelligence using optical coherence tomography (OCT), The OCT imaging system head is positioned in close proximity to a biological sample containing oocytes, wherein the OCT is operably connected to an artificial intelligence / machine learning system (AI / ML system) and an imaging system, and the imaging system includes a camera system and a lighting system. To create at least one three-dimensional image of the oocyte using the OCT, AI / ML system and imaging system, and Analyzing the three-dimensional image using the AI / ML system, wherein the analysis includes detecting the presence or absence of polar bodies based at least partially on the planar view of the oocyte. Methods that include...

142. The method according to claim 141, wherein the oocyte is a plurality of oocytes.

143. The method according to claim 141, wherein the oocyte is an embryo.

144. The method according to claim 141, wherein the embryo is a plurality of embryos.

145. The method according to claim 141, wherein the OCT is a polarization-sensitive OCT.

146. The method according to claim 141, wherein the OCT is positioned above the biological sample.

147. The method according to claim 141, wherein the OCT is positioned below the biological sample.

148. The method according to claim 141, wherein the OCT is positioned on the side surface of the biological sample.

149. The method according to claim 141, wherein the OCT can move to multiple locations relative to the biological sample and generate images from those locations.

150. The method according to claim 141, wherein the lighting system includes polarized lighting.

151. The method according to claim 141, wherein the three-dimensional image is a fusion of a plurality of images combined with the three-dimensional image.

152. The method according to claim 151, wherein the image fusion body includes image components from the OCT and at least one other microscope system.

153. The method according to claim 141, wherein the three-dimensional image includes a simulated virtual element.

154. The method according to claim 141, wherein the AI / ML and imaging system automatically identifies the presence or absence of a meiotic spindle based at least partially on the three-dimensional image.

155. The method according to claim 154, wherein the AI / ML and imaging system automatically evaluate the meiotic spindle, form a prediction algorithm, and determine the probability of whether the oocyte will mature sufficiently.

156. The method according to claim 154, wherein the AI / ML and imaging system automatically evaluates the meiotic spindle and forms a prediction algorithm to determine the probability of whether the oocyte will mature in further incubation.

157. The method according to claim 141, wherein the AI / ML and imaging system automatically determines the best arrangement of the oocytes for use during injection.

158. The method according to claim 141, wherein the AI / ML and imaging system automatically evaluates the film integrity.

159. A method for oocyte identification and maturity assessment based on automated artificial intelligence using optical coherence tomography (OCT), The method involves positioning an OCT imaging system head on the side of a biological sample containing oocytes, wherein the OCT is operably connected to an artificial intelligence / machine learning system (AI / ML system) and an imaging system, and the imaging system includes a camera system and a lighting system. Using polarized light directed at the biological sample containing the oocytes, To create at least one three-dimensional image of the oocyte using the OCT, AI / ML system and imaging system, Analyzing the three-dimensional image using the AI / ML system, wherein the analysis includes detecting the presence or absence of a meiotic spindle based at least partially on the three-dimensional image, and The maturity of the oocyte is evaluated at least partially based on the detection of the presence or absence of the meiotic spindle. Methods that include...

160. The method according to claim 159, wherein the three-dimensional image includes a simulated virtual element.

161. A method for IVF microtool control based on automated artificial intelligence, A control device associated with an artificial intelligence / machine learning system (AI / ML system) and an imaging system, which is used at least partially to operate the robotic components of an in vitro fertilization (IVF) module, receives commands. To retrieve at least one IVF microtool assembly (MA) from a tool stock location using a first robotic mechanism of the robotic system, at least partially based on the command, To retrieve at least one IVF receptor using a second robotic mechanism of the robotic system, at least partially based on the command, The IVF receptor is positioned on a stage within the IVF module, at least partially based on the command. The at least one MA is positioned in close proximity to the IVF receptor using the first robotic mechanism, at least partially based on the command, wherein the positioning is in accordance with a planned IVF procedure, and its operation is at least partially controlled by the AI / ML system and the imaging system. To confirm the correct placement of the at least one MA adjacent to the IVF receptor within the IVF module, and Send a warning to confirm the correct placement. Methods that include...

162. The method according to claim 161, wherein the first robot mechanism and the second robot mechanism are a single robot mechanism.

163. The method according to claim 161, wherein the first robot mechanism is a plurality of robot mechanisms.

164. The method according to claim 161, wherein the second robot mechanism is a plurality of robot mechanisms.

165. The method according to claim 161, wherein the recovery of the IVF receptor is from a transport vehicle used to hold or move the IVF receptor.

166. The method according to claim 165, wherein the transport mechanism is a belt conveyor.

167. The method according to claim 165, wherein the transport mechanism is a rail conveyor.

168. The method according to claim 165, wherein the transport device includes at least one form capable of holding the IVF receptor.

169. The method according to claim 165, wherein the transport mechanism includes a temperature control plate located in close proximity to the IVF receptor in order to maintain a predetermined temperature of the biological sample.

170. The method according to claim 161, wherein the imaging system includes a microscope system and a camera system and an illumination system.

171. The method according to claim 161, wherein the inventory location on the stage is a plurality of locations.

172. The method according to claim 161, wherein the IVF receptor is a dish.

173. The method according to claim 161, wherein the IVF receptor is a tube.

174. The method according to claim 161, wherein the IVF receptor is a microtool.

175. The method according to claim 161, wherein the IVF receptor is an ICSI needle.

176. A method for embryo manipulation based on automated artificial intelligence, The method involves positioning an in vitro fertilization (IVF) receptor containing an embryo on a transport mechanism within an oocyte preparation module using a microtool assembly (MA) controlled by a first robotic mechanism, wherein the transport mechanism includes at least one form capable of holding the IVF receptor and a temperature control plate adjacent to the IVF receptor to maintain a predetermined temperature for the IVF receptor and the embryo, and the positioning is at least partially performed using an artificial intelligence / machine learning system (AI / ML system) and an imaging system, wherein the imaging system includes a microscope system, a camera system and an illumination system. The IVF receptor is transported from the oocyte preparation module to the incubation module using the transport mechanism, wherein the transport of the IVF receptor is performed to a predetermined location within the incubator using the AI / ML system and the imaging system. Transporting the IVF receptor from the incubation module to a position proximal to the microscope stage after a predetermined period of time, wherein the position is at least partially determined by the AI / ML system and the imaging system. To evaluate the developmental stage of the embryo, and Transporting the embryos to the IVF platform location based at least partially on maturity assessment. Methods that include...

177. The method according to claim 176, wherein the evaluation of the embryonic developmental stage includes imaging for detecting the presence of polar bodies.

178. The method according to claim 176, wherein the embryonic developmental stage evaluation is a developmental evaluation performed on multiple embryos within a cohort.

179. The method according to claim 178, wherein the cohort is defined in part on the basis of embryos sharing a common IVF receptor.

180. The method according to claim 176, wherein the AI / ML system generates a maturity score based at least in part on the developmental stage evaluation.

Images