Method implemented by computer system and oocyte mass analysis system
By extracting the morphological and mechanical characteristics of oocytes, and evaluating the quality index of oocytes using machine learning models, the accuracy and automation of selecting high-quality oocytes in the existing technology are solved, and more efficient embryo selection is achieved.
Patent Information
- Application Number
- CN202380017854.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-03-24
- Filing Date
- 2023-07-18
- Publication Date
- 2024-11-26
AI Technical Summary
Existing embryo selection techniques rely on manually adjusted models and grading systems, are prone to errors and cannot effectively utilize sensor information, resulting in inaccurate accuracy and automation technology of embryo selection.
By obtaining the image sequence of oocytes and sensor information of applied pressure, extracting morphology and mechanical characteristics, using machine learning models (such as support vector machines and neural networks) to determine the quality indicators of oocytes, and providing a standard grading system to select high-quality oocytes. cell.
A more accurate and objective assessment of oocyte quality is achieved, the accuracy of selecting high-quality oocytes is improved, the artificial subjectivity is reduced, and the efficiency of automated analysis is enhanced.
Smart Images

Figure CN119032377A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to a medical analysis using a machine learning model, and to an analysis of oocytes using sensor information and a machine learning model. Background Art
[0002] In recent years, treatments for infertility have advanced. An exemplary treatment includes in vitro fertilization ("IVF"). IVF begins with an ovarian stimulation phase to stimulate egg production. Eggs (oocytes) can be removed from the patient and fertilized in vitro to form embryos. The embryos can be subjected to a number of tests and analyses in an effort to select the most viable and / or advantageous embryos for implantation. However, there are technical issues with such tests and analyses, and therefore, there are technical challenges with the accurate protocols for achieving this selection.
[0003] For example, different grading systems have been developed to assist in determining the viability of each embryo. These grading systems are generally related to the manual annotation of embryo images or time-lapse dynamic images. It is understandable that, based on the subjective judgment of, for example, embryologists, the selection process has been shown to be fallible. In addition, there is no currently commonly used standard grading system to select high-quality oocytes. The automation technology currently utilized for embryo analysis is inaccurate, and different sensor information cannot be utilized. Summary of the invention
[0004] The present invention provides a method implemented by a computer system and an oocyte quality analysis system, which has accuracy, can select high-quality oocytes using a standard grading system, and utilizes different sensor information.
[0005] The method of the present invention implemented by a system of one or more computers includes the following steps. A plurality of images forming an image sequence associated with a time period are acquired, the image sequence depicting an oocyte and a portion of a tool for applying pressure to the oocyte, wherein each image is associated with a respective pressure value applied to the oocyte at a respective time of image capture. Thereafter, based on the images and the pressure values, features associated with the oocyte are determined, the features including morphological features representing measurements of the oocyte within the time period and mechanical features representing deformation of the oocyte within the time period. Next, one or more indicators representing oocyte quality are determined via a machine learning model based on input values including at least a subset of features, wherein a specific indicator represents blastocyst formation. The one or more indicators are configured to be presented via an interactive user interface.
[0006] In one embodiment of the invention, the oocytes depicted in the images are mammalian.
[0007] In one embodiment of the invention, the image depicts an oocyte being sucked into a portion of the tool.
[0008] In one embodiment of the present invention, the mechanical characteristic represents elasticity of the oocyte, and wherein the mechanical characteristic is determined based on a suction depth of the oocyte into the portion of the tool.
[0009] In one embodiment of the present invention, the mechanical characteristic represents the viscoelastic behavior of the oocyte.
[0010] In one embodiment of the invention, the morphological feature indicates at least the depth of aspiration of the oocyte of the corresponding image into the tool portion.
[0011] In one embodiment of the present invention, one of the morphological features is derived by normalizing the aspiration depth of the corresponding imaged oocyte into the tool portion by the inner diameter of the tool.
[0012] In one embodiment of the present invention, the morphological feature represents a first measurement associated with the cytoplasm of the oocyte and a second measurement associated with the zona pellucida of the oocyte.
[0013] In an embodiment of the invention, the tool portion is a micropipette, and wherein the micropipette is adjacent to the oocyte during the period of time and is configured to apply pressure.
[0014] In an embodiment of the invention, a first image associated with the beginning of the time period depicts the oocyte prior to application of pressure, and wherein a second image subsequent to the first image in the sequence of images depicts the oocyte with pressure applied.
[0015] In one embodiment of the present invention, the pressure value is selected from a range of pressures applied within a time period.
[0016] In one embodiment of the present invention, the machine learning model is a support vector machine.
[0017] In one embodiment of the present invention, the machine learning model is a neural network.
[0018] In one embodiment of the invention, the method implemented by the system of one or more computers further comprises the following steps: Calculate the measured value of the pressure applied to the oocyte based on the pressure value and the geometric information related to the tool part. Obtain the mechanical characteristics by fitting the suction depth curve using a linear elastic method, wherein the fitted suction depth curve is based on the measured values of the suction depth and pressure of the oocyte entering the pressure tool in the corresponding image.
[0019] In one embodiment of the present invention, the input value further includes clinical information of the person related to the oocyte, and the clinical information includes one or more of age, body mass index, developmental stage related to the oocyte, and cryopreservation information.
[0020] The oocyte quality analysis system of the present invention includes one or more processors and a non-transitory computer storage medium storing instructions. When executed by one or more processors, the one or more processors:
[0021] acquiring a plurality of images forming an image sequence associated with a time period, the image sequence depicting an oocyte and a portion of a tool applying pressure to the oocyte, wherein each image is associated with a respective value of pressure applied to the oocyte at a respective time of image capture;
[0022] Determining features associated with the oocyte based on the image and the pressure values, the features comprising morphological features representing measurements of the oocyte over the time period and mechanical features representing deformation of the oocyte over the time period; and
[0023] determining, via a machine learning model based on input values comprising at least a subset of features, one or more indicators indicative of oocyte quality, wherein a particular indicator is indicative of blastocyst formation,
[0024] One or more of the indicators are configured to be presented via an interactive user interface.
[0025] In one embodiment of the invention, the image depicts an oocyte being sucked into a portion of the tool.
[0026] In one embodiment of the present invention, the mechanical characteristic represents elasticity of the oocyte, and wherein the mechanical characteristic is determined based on a suction depth of the oocyte into the portion of the tool.
[0027] In one embodiment of the present invention, the mechanical characteristic represents the viscoelastic behavior of the oocyte.
[0028] In one embodiment of the invention, the morphological feature is indicative of at least the depth of aspiration of the oocyte into the tool portion of the corresponding image.
[0029] In one embodiment of the present invention, the morphological feature represents a first measurement associated with the cytoplasm of the oocyte and a second measurement associated with the zona pellucida of the oocyte.
[0030] In one embodiment of the present invention, the pressure value is selected from a range of pressures applied within a time period.
[0031] In one embodiment of the invention, the instructions further cause one or more processors to:
[0032] calculating a measure of the pressure exerted on the oocyte based on the pressure value and geometric information associated with the tool portion; and
[0033] The mechanical characteristics are obtained by fitting an aspiration depth curve using a linear elastic method, wherein the fitted aspiration depth curve is based on the measured values of the aspiration depth and pressure of the oocyte entering the pressure tool in the corresponding image.
[0034] In one embodiment of the present invention, the input value further includes clinical information of the person related to the oocyte, and the clinical information includes one or more of age, body mass index, developmental stage related to the oocyte, and cryopreservation information.
[0035] The method of the present invention implemented by a system of one or more computers comprises the following steps. A plurality of image sequences depicting corresponding oocytes are obtained, wherein each image sequence depicts each oocyte and a portion of a tool for applying pressure to the oocyte, and wherein pressure values associated with images included in the image sequence are obtained. Thereafter, one or more indicators representing the oocyte quality of each oocyte are determined by a machine learning model, wherein a specific indicator represents blastocyst formation, and wherein the input value of the machine learning model for each oocyte includes a morphological feature representing the measurement of the oocyte and a mechanical feature representing the deformation of the oocyte during the application of pressure to the oocyte. Next, information representing the selection of one or more oocytes is output based on the indicator.
[0036] In one embodiment of the invention, the image depicts an oocyte being sucked into a portion of the tool.
[0037] In one embodiment of the present invention, the mechanical characteristic represents elasticity of the oocyte, and wherein the mechanical characteristic is determined based on a suction depth of the oocyte into the portion of the tool.
[0038] In one embodiment of the present invention, the morphological feature represents a first measurement associated with the cytoplasm of the oocyte and a second measurement associated with the zona pellucida of the oocyte.
[0039] In one embodiment of the present invention, the information representing the selection is output via an interactive user interface, and wherein the interactive user interface:
[0040] Presenting a graphic depiction of one or more oocytes,
[0041] selecting a particular oocyte from among the one or more oocytes in response to user input; and
[0042] Presents detailed information related to a specific oocyte.
[0043] In an embodiment of the invention, information representative of the selection is output via an interactive user interface, and wherein the interactive user interface presents respective text descriptions of respective oocytes summarizing the one or more oocytes.
[0044] The present invention provides a method and an oocyte quality analysis system implemented by a computer system, which is accurate and can select high-quality oocytes using a standard grading system and utilize different sensor information. In this way, it can be effectively used to select the most viable and / or most advantageous embryos for implantation, so as to be applied in the field of infertility treatment. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 is a block diagram of an example oocyte analysis system according to some embodiments of the present disclosure, the system including a feature pre-processing engine and a machine learning model.
[0046] Figure 2A Description by Figure 1 Example morphological features of an oocyte generated by the feature preprocessing engine.
[0047] Figure 2B illustrate Figure 1 Mechanical features of oocytes generated by the feature preprocessing engine.
[0048] Figure 3A An example user interface including example processing results is described.
[0049] Figure 3B Describe the example user interface, Figure 1 The example oocyte analysis system receives user input through the user interface.
[0050] Figure 3C Another example user interface for presenting processing results is described.
[0051] Figure 4 is a flow chart of an example process for determining indicators representative of oocyte quality.
[0052] Figure 5 is a flow chart of an example process for selecting a subset of a plurality of oocytes based on a metric indicative of quality.
[0053] Figure 6 The overall architecture of an example oocyte analysis system according to some embodiments of the present disclosure is described. DETAILED DESCRIPTION
[0054] This specification describes a technique for selecting high-quality eggs (e.g., oocytes), for example, these eggs have a higher probability of blastocyst formation. As will be described, the system can use different sensor information to analyze oocytes using machine learning techniques. The sensor information may, for example, include an image of an oocyte (unfertilized egg) that undergoes deformation due to applied pressure and a pressure measurement representing the pressure applied to the oocyte. When the oocyte is deformed, this sensor information can help understand the morphology and mechanical properties of the oocyte, so that the quality of each oocyte can be more accurately assessed. In contrast, previous technologies rely on manually adjusted models or grading systems that use a single type of information (e.g., images) and are prone to errors. As will be described, the disclosed technology uses different sensor information, together with a specially trained model, to more accurately and effectively assess the quality or viability of eggs in the early stages.
[0055] To assess egg viability, some prior techniques rely on visual assessment of embryos (e.g., fertilized eggs) by embryologists. Some clinics record images of embryos, and embryologists can score embryos based on various grading systems and their visual assessments. A major challenge with embryo selection is the high level of labor, subjectivity, and variability between embryologists of varying skill levels and grading systems that perform differently. Specifically, embryologists, after spending considerable time visually assessing embryos, often disagree with each other, and even with themselves, about which embryo has the best survival for transplantation. Furthermore, it is unclear which embryo characteristics associated with a particular grading system ultimately predict the success rate of each embryo.
[0056] Other prior techniques include automated techniques for selecting high-quality eggs. These techniques may rely on specific characteristics of the fertilized eggs. Typically, the specific characteristics of the fertilized eggs are obtained by analyzing (e.g., using microscopy and computer vision techniques) dynamic images of the changes in the eggs during growth and development. However, selecting eggs based on these characteristics (e.g., characteristics derived from observing the growth and development of eggs) may not produce satisfactory results.
[0057] In contrast, the disclosed techniques allow for analysis of egg viability at an earlier stage (e.g., when the egg is still unfertilized rather than when the egg is fertilized). Thus, the disclosed techniques can eliminate the additional complexity associated with fertilized eggs that may not later implant. By utilizing a machine learning model to select oocytes based on extracted oocyte features, the disclosed techniques can obtain a more objective and quantitative analysis of oocyte viability.
[0058] In addition, the disclosed technology utilizes machine learning techniques to analyze the morphology and mechanical characteristics of the oocyte. As will be described, the morphological characteristics may include geometric information related to the oocyte, such as the size or length of the zona pellucida, cytoplasm, polar bodies, perivitelline space, the degree to which the oocyte is sucked into the pressure application tool (e.g., suction depth), etc. As will be described, the mechanical characteristics may include parameters determined or derived based on the deformation characteristics of the oocyte. For example, the mechanical characteristics may include at least morphodynamic parameters described in a specific model (e.g., a Zener model), etc.
[0059] These machine learning techniques can utilize features to represent one or more indicators that represent oocyte quality. Example indicators may include a value representing the likelihood or probability of blastocyst formation. Additional example indicators may include an indication of oocyte "goodness", the likelihood or probability of aneuploidy, and implantation rate, etc.
[0060] More specifically, the system can obtain an image sequence depicting an oocyte that is deformed based on a mechanical stimulus. An exemplary stimulus can include aspirating the oocyte into a portion of a pressure tool (e.g., a micropipette) that applies pressure to the oocyte. The system can process the image sequence using example computer vision techniques to derive the above-mentioned morphological features and mechanical features. In some examples, in order to process the different resolutions of the hardware (e.g., a microscope camera) used to capture the image, a normalization technique can be used. The normalization technique can be integrated into a part of an image processing technique performed on a captured image sequence associated with the oocyte. The normalization technique will be described in more detail below.
[0061] The extracted morphological and mechanical features of the oocytes can then be used to train a machine learning model or to perform inference using the machine learning model. The training process can include training the machine learning model using a subset of the morphological and mechanical features. The trained machine learning model can then be used to generate one or more indicators representing oocyte quality, wherein a specific indicator can represent the formation of a blastocyst from an oocyte.
[0062] In addition, optionally, the pointer can be presented to the user or professional (e.g., embryologist) through an interactive user interface. Therefore, the viability of the oocyte can be further analyzed or evaluated more effectively. Therefore, based on the embodiments of the present disclosure, a more objective, automated and time-saving evaluation of the viability of the oocyte can be achieved.
[0063] For example, the accuracy associated with the disclosed technology, a total of 185 oocytes were evaluated to assess their ability to form blastocysts. The 185 samples were divided into 80% for training the machine learning model and 20% for testing the machine learning model. The machine learning used for viability assessment is a support vector machine (SVM). The machine learning model is trained to predict whether a specific oocyte in a sample will have a blastocyst. The prediction results were statistically analyzed, showing an accuracy of 73%, a sensitivity of 85%, a specificity of 59%, a positive predictive value (PPV) of 77%, and a negative predictive value (NPV) of 71%. Compared to the statistics compiled based on the predictions made by embryologists, showing an accuracy of 45%, a sensitivity of 59%, a specificity of 33%, a PPV of 43%, and a NPV of 49%, the system and method according to the present disclosure statistically achieved better performance.
[0064] The foregoing aspects and many of the attendant advantages of the present disclosure will become more readily appreciated as they become better understood when reference is made to the following description taken in conjunction with the accompanying drawings.
[0065] Example Block Diagram
[0066] Figure 1 1 is a block diagram of an example oocyte analysis system 100 according to some embodiments of the present disclosure, which includes a feature preprocessing engine 120 and a machine learning model 130. As shown, the feature preprocessing engine 120 can receive as input an image sequence 102 of an oocyte (e.g., a mammalian oocyte, such as a human oocyte), a pressure value 106, and clinical information 108. The feature preprocessing engine 120 can then generate as output features 122 associated with the oocyte. The features 122 can be input into the machine learning model 130, which outputs oocyte quality information 132. The oocyte quality information 132 can be presented to a user for further analysis.
[0067] The image sequence 102 may include a plurality of images (e.g., images 104A to 104N), each of which depicts an oocyte 110 together with a pressure tool 112. More specifically, the images 104A to 104N may form a dynamic image depicting the process of aspirating the oocyte 110 into the pressure tool 112. Image 104A may, for example, represent the first frame of the dynamic image, while image 104N may, for example, represent the last frame of the dynamic image. Image 104A depicts that the oocyte 110 has not yet been aspirated into the pressure tool 112, while image 104N depicts that at least a portion of the oocyte 110 has been aspirated into the pressure tool 112. In some examples, the image sequence 102 may have a frame rate of 10 Hz, 20 Hz, 70 Hz, 3000 Hz, and the total duration of the dynamic image is 1 second, 2 seconds, 10 seconds, etc.
[0068] In addition, in some embodiments, the pressure tool 112 is a micropipette having a diameter between certain thresholds (e.g., between 10 micrometers (μm), 20 μm, 40 μm and 60 μm, 70 μm, 100 μm, etc.). The micropipette can apply negative pressure on the oocyte 110 (e.g., the pressure inside the micropipette is lower relative to the pressure outside the micropipette) to draw the oocyte 110 into the micropipette without damaging the oocyte 110. Exemplary pressures may include -0.01 psi to -0.5 psi. For example, the micropipette may abut or otherwise contact the oocyte. Although the image sequence 102 illustrates that the oocyte 110 is drawn into the pressure tool 112, in some embodiments, the pressure tool 112 may apply other forms of mechanical stimulation (e.g., positive pressure). In this way, different morphological responses of the oocyte 110 may be obtained. The feature preprocessing engine 120 may use the morphological responses to analyze and / or extract different morphological features.
[0069] exist Figure 1 , a feature pre-processing engine 120 is shown using the image sequence 102, the pressure values 106, and the clinical information 108 to output a feature 122 associated with the oocyte 110. In some embodiments, all of this information may be used to determine the feature 122. In some embodiments, a subset of the information may be used.
[0070] Regarding the pressure value 106, the pressure value 106 may include a plurality of numerical values that indicate how much force is applied to the oocyte 110 during a time range or time period of the images 104A to 104N. For example, the pressure value 106 may indicate that at the time (e.g., time or timestamp) when the image 104A was captured, a first pressure (e.g., -0.3 psi) was applied to the oocyte 110, and at the time when the image 104N was captured, an Nth pressure was applied to the oocyte 110. The pressure values applied to the oocyte 110 may be the same or may be different at different images in the sequence of images 102.
[0071] In some examples, the pressure applied to the oocyte 110 may increase over time, while in other examples, the pressure applied may decrease over time. In addition, the pressure value 106 may include a force applied to the oocyte 110, wherein the force may be calculated based on the pressure generated by the pressure tool 112 and applied to the oocyte 110. The force applied to the oocyte 110 may be used to derive a mechanical characteristic of the oocyte 110, which will be described below with reference to Figure 2B Describe it in more detail.
[0072] The clinical information 108 may include the age and body mass index (BMI) of the patient from whom the oocyte 110 was derived. In addition, the clinical information 108 may include information indicating whether the oocyte 110 has been cryopreserved (CP). The clinical information 108 may also indicate the available number of mature oocytes (MII) associated with the patient.
[0073] Based on at least some of the image sequences 102, the pressure values 106, and the clinical information 108, the feature preprocessing engine 120 can extract features 122 associated with the oocyte 110. The features 122 can include morphological features (size or length of the zona pellucida, cytoplasm, polar bodies, or perivitelline space) and mechanical features (e.g., elasticity and / or viscosity) of the oocyte 110. Certain features can be generated based on each image in the image sequence 102, while other features can be determined based on all images or a subset of the images. Figure 2A-2B These features 122 are described in more detail.
[0074] Based on the features 122, the machine learning model 130 may generate oocyte quality information 132 for the oocyte 110. Example information 132 may include the possibility of blastocyst formation. In some embodiments, the machine learning model 130 may be a support vector machine (SVM) trained to output the information 132. In other examples, the machine learning model 130 may be a deep learning model. For example, the deep learning model may include a recurrent neural network (RNN) trained to output the information 132. In the example, the RNN may input the features 122 as a sequence, and may output the information 132 of the sequence. The model may also be a convolutional neural network or a fully connected network. The machine learning model 130 may use all or a subset of the features 122 to determine the oocyte quality information 132. For example, the SVM may be trained to use a subset of the features 122.
[0075] In some embodiments, the oocyte quality information 132 may also indicate the likelihood that the oocyte 110 will form a "usable" blastocyst, where "usable" means that the blastocyst formed from the oocyte 110 is suitable for transfer or for implantation. Additionally, or alternatively, the oocyte quality information 132 may indicate the likelihood that the blastocyst formed from the oocyte 110 will be an "unusable" blastocyst, where "unusable" means that the blastocyst formed from the oocyte 110 will be in such a poor state that the blastocyst is not suitable for further IVF treatment.
[0076] Block Diagram-Morphological Feature Generation
[0077] Figure 2A FIG20 shows example morphological features of an oocyte generated by the feature pre-processing engine 120. The example morphological features 210 may form Figure 1Example morphological features 210 may include aspiration depth of the oocyte 110, diameter of the pressure tool 112 (e.g., a micropipette), size and / or shape of the cytoplasm and zona pellucida of the oocyte 110, etc. Optionally, these features may be determined for each input image.
[0078] The determination of some example morphological features 210 of the oocyte 110 may be illustrated by oocyte measurements 202N, which depict example measurements (e.g., bounding box A, distance B, distance C, distance D, distance E, bounding box F, and distance G). Oocyte measurements 202N correspond to measurements of image 104N, while oocyte measurements 202A correspond to measurements of image 104A.
[0079] With respect to the illustrated oocyte measurement 202N, bounding box A may represent a region of interest (ROI) that is clipped for performing further image processing on the image sequence 102. Distance B may represent the length and / or size of the upper side of the zona pellucida of the oocyte 110. Distance C may represent the length and / or size of the right side of the zona pellucida of the oocyte 110. Distance D may represent the length and / or size of the lower side of the zona pellucida of the oocyte 110. Distance E may represent the inner zona pellucida of the oocyte 110. Distance G may represent the length of the inner diameter of the pressure tool 112. Bounding box F may represent a region of interest (ROI) that may be clipped for calculating the suction depth of the oocyte 110 into the pressure tool 112. Calculating the suction depth will be described in more detail below.
[0080] Different image processing and object recognition techniques can be used to obtain the example morphological features 210. In some examples, computer vision techniques can be used to derive the example morphological features 210. For example, edge detection techniques can be used to identify boundaries associated with different parts of an oocyte. In other examples, an image segmentation model based on a deep learning algorithm can be used to derive the example morphological features 210. For example, an image segmentation mask can be generated for the feature 210. It is understood that the image segmentation mask can assign a color or pixel value, indicating whether the pixel forms part of a classification feature (e.g., a zona pellucida, etc.). The derivation of each example morphological feature will be discussed in more detail below.
[0081] Bounding box A can define an ROI that will be cropped for any image (e.g., image 104A or image 104N) of the image sequence 102, wherein some morphological features 210, such as the inner boundary of the zona pellucida, can be used to calculate the size of the pressure tool 112 and the suction depth of the oocyte 110 within this ROI. In some examples, the bounding box A can be identified according to the following example techniques. For example, the tip of the pressure tool 112 (e.g., a micropipette) contacting the oocyte 110 can be identified and / or located based on a micropipette tip image that can be acquired in advance. Based on the position of the micropipette tip, the bounding box A can be drawn to determine the ROI around the micropipette tip. Although the bounding box A is shown as a rectangle, other shapes can be used to define the ROI.
[0082] Distance B, distance C and distance D define the length or thickness of the upper, right and lower zona pellucida, respectively. By averaging the lengths of these distances, a morphological feature representing the average size of the zona pellucida of the oocyte 110 can be obtained. In addition, distance E can be defined based on the position of the micropipette tip obtained when delimiting the inner boundary of the zona pellucida of the oocyte 110 and the bounding box A above.
[0083] As an example, distance E can be used to derive the horizontal (e.g., in the direction of arrow C and arrow E) length of the cytoplasm of the oocyte 110. More specifically, the horizontal length of the cytoplasm of the oocyte 110 can be derived by subtracting the combined length of arrow E and arrow C from the horizontal length of the oocyte 110 (e.g., defined by the right end of arrow C and the left end of arrow E). The horizontal length of the cytoplasm of the oocyte 110 can also be used as part of the example morphological feature 210. Similarly, the vertical (e.g., in the direction of arrow B and arrow D) length of the cytoplasm of the oocyte 110 can be derived by subtracting the combined length of arrow B and arrow D from the vertical length of the oocyte 110 (e.g., defined by the top end of arrow B and the bottom end of arrow D). The vertical length of the cytoplasm of the oocyte 110 can also be used as part of the feature 122.
[0084] The bounding box F can be used as a ROI to calculate the suction depth of the oocyte 110 (e.g., how much longer the oocyte 110 is sucked into the pressure tool 112 compared to the position of the oocyte 110 just before being sucked). The suction depth may be helpful in deriving mechanical characteristics associated with the oocyte 110, which will be referred to below. Figure 2B For more detailed discussion. In some embodiments, the suction depth can be derived by determining the deepest horizontal position (e.g., the end of the suction depth) at which the oocyte 110 is sucked into the pressure tool 112. The distance between the end of the suction depth and the inner boundary of the zona pellucida of the oocyte 110 can then be calculated. The following is an example technique for determining the suction depth.
[0085] The ROI (e.g., bounding box F) can be delimited (e.g., identified) based on the position of the tip of the pressure tool 112 in contact with the oocyte 110. As shown in oocyte measurement 202N, bounding box F is rectangular; however, in other examples, bounding box F can have different shapes (e.g., square or irregular shapes). In some embodiments, bounding box F extends to the leftmost portion of the image at a specific threshold distance. That is, the distance can be set to ensure that it contains the terminal end or end of the oocyte sucked into the tool 112.
[0086] Pixels that are vertically aligned with each other within the bounding box F may be summed to derive a pixel curve. More specifically, assuming that the bounding box F spans N pixels horizontally and M pixels vertically, the pixel intensities of pixels having the same horizontal position within the bounding box F may be summed. For example, M pixels may be summed at a particular horizontal position. Thus, N may be a positive integer that depends on the resolution of the image sequence 102 and the horizontal length of the bounding box F. For example, if there are 50 horizontal positions, then for each position, the pixels extending vertically within the bounding box F may be summed to derive a value.
[0087] These values can then be used to generate a pixel curve having a sum value on a first axis (e.g., y-axis) and a number of horizontal positions on a second axis (e.g., x-axis). The pixel curve can be smoothed using a signal processing technique such as a sliding window moving average.
[0088] Subsequently, a first order derivative may be determined on the smoothed pixel curve. The derivative may represent the vertically added intensity. The minimum of the derivative of the pixel curve may then be found, indicating the end of the aspiration depth of the oocyte 110 for a particular image (e.g., image 104A).
[0089] The above-mentioned example techniques can be performed on other images in the image sequence 102, so that the end point of each suction depth can be obtained for images 104A to 104N. In some examples, the suction depth can be calculated based on the end of the suction depth and the inner boundary of the zona pellucida of the oocyte 110. By calculating the suction depth of each image in the images 104A to 104N in the image sequence 102, over time, the suction depth of the oocyte 110 can be plotted as a chart, wherein the first axis shows time and the second axis shows the suction depth of the oocyte 110. In some examples, the suction depth can be further normalized according to the size of the zona pellucida of the oocyte 110. For example, the suction depth can be normalized by deducting the mean value of distance B, distance C and distance D from the calculated suction depth.
[0090] Although the examples of determining the suction depth are described above, it is understood that other techniques may be used and fall within the disclosure of this article. For example, deep learning techniques may be used to determine the boundary or end of the oocyte entering the tool. As another example, edge detection techniques (e.g., edge detection kernels) may be used to identify the boundary of the oocyte in the tool.
[0091] In addition, optionally, feature preprocessing engine 120 may not need to analyze all images in image sequence 102. In some examples, by selecting a starting frame and performing analysis on the frames after the starting frame and the starting frame, only a portion of image sequence 102 is analyzed. The starting frame can be a frame that captures the image of oocyte 110 just before oocyte 110 is sucked into pressure tool 112. For example, the system can identify a frame associated with a timestamp before the time or a timestamp associated with applying negative pressure. As another example, the system can analyze image frames and identify a frame before a frame with a suction depth greater than a threshold value (e.g., zero, a small value, etc.). Advantageously, analysis can be more time-saving and consume less computing power.
[0092] In addition, optionally, the pre-processing engine 120 can reduce (e.g., reduce) the number of frames to be analyzed by removing the continuous frames in which the oocyte 110 remains stationary or motionless. For example, the pre-processing engine 120 can determine that only 0.5 seconds of the dynamic image depicted by the image sequence 102 shows the movement of the oocyte 110, while the remaining 1.5 seconds of the dynamic image of the image sequence 102 shows that the oocyte 110 remains motionless or remains relatively still. The pre-processing engine 120 can then extract 0.5 seconds of the dynamic image for further analysis. Advantageously, the analysis time can be reduced by shortening the length of the dynamic image used for analysis.
[0093] The feature pre-processing engine 120 can additionally determine the inner diameter of the pressure tool 112, which is represented by the distance G in the oocyte measurement 202N. In some embodiments, the length of the inner diameter of the pressure tool 112 can be obtained by counting the number of pixels vertically aligned along the inner diameter of the pressure tool 112. The inner diameter of the pressure tool 112 can be used as a normalization factor to increase the interoperability of the oocyte analysis system 100 across different dynamic image capture platforms that may have different hardware specifications (e.g., resolution of captured images).
[0094] Normalization may also be related to the z-height of the oocyte in the object that holds the oocyte when the image is taken. That is, the oocyte may be placed in water or another liquid and change height (e.g., closer to the bottom or top of the water). Therefore, this height adjustment may change the size of the oocyte compared to another oocyte in a different image in the sequence 102 or to the same oocyte. For example, the height of the oocyte may change slightly during the image sequence 102.
[0095] More specifically, over time, the suction depth of the oocyte can be normalized by the ratio between the inner diameter of the pressure tool 112 and the pixel length of a particular image sequence. Using computer vision techniques, the inner diameter of the pressure tool 112 can be identified, and then the length of the inner diameter can be calculated by calculating the number of pixels along the axis covered by the inner diameter of the pressure tool 112. For example, the inner diameter of the pressure tool 112 can span 1,000 pixels, and then the suction depth can be divided by 1,000 for normalization. By normalizing the suction depth of the oocyte by the inner diameter of the pressure tool 112, the oocytes of different frames acquired under different image capture settings can be placed on the same basis for calculating morphological features associated with the oocyte. Therefore, the interoperability and reproducibility of the currently disclosed system and method can be enhanced.
[0096] The feature preprocessing engine 120 may also acquire additional morphological features associated with the oocyte 110. Exemplary features may include the size and / or length of the polar body, cytoplasm, and / or perivitelline space (e.g., PVS, the space between the zona pellucida and the cytoplasm) of the oocyte 110. An image segmentation model based on deep learning algorithms and computer vision techniques may be used to determine these additional morphological features. Advantageously, the additional available morphological features may be potentially useful for training and / or testing the machine learning model 130 to improve the accuracy, sensitivity, specificity, NPV, and PPV of the results generated by the machine learning model 130.
[0097] Blocks Figure 1 Mechanical feature generation
[0098] Figure 2B The diagram shows Figure 1 The mechanical features of the oocyte generated by the feature preprocessing engine 120 are as follows. Figure 2B As shown, the feature pre-processing engine 120 receives the image sequence 102 and the pressure value 106 as input and generates an example mechanical feature 254 based on the example elastic model 252. The example mechanical feature 254 may represent deformation or movement of the oocyte 110 during the image sequence 102.
[0099] The pressure value 106 may include the force applied to the oocyte 110 (e.g., by the pressure tool 112 described above). Specifically, the force applied to the oocyte 110 may be calculated based on the following equation (A). Thus, the force may be based on the pressure value and geometric information associated with the micropipette.
[0100]
[0101] in
[0102] The unit of pressure is psi
[0103] psi = ratio of Pa to Psi
[0104] · Product = (d / 2) 2 *pi,
[0105] d = inner diameter of the micropipette
[0106] μ = ratio of m to μm
[0107] π = the ratio of a circle’s circumference to its diameter (A)
[0108] The feature preprocessing engine 120 may optionally combine the above-mentioned forces Figure 2A Feature 254 is determined by the inhalation depth described in .
[0109] In some embodiments, a linear elastic model, a modified linear elastic model, or a standard linear solid model may be used to obtain characteristics indicative of the viscoelastic behavior of the oocyte 110. Example models may include a Zener model, a modified Zener model, and the like. For example, the following equation (B) may be used to fit the inhalation depth data points of the oocyte 110 over time to obtain the mechanical characteristics of the oocyte 110. For example, Figure 2B As shown, the model described above (e.g., spring-damper model) can be used with two springs and two dampers to simulate mechanical properties. The model can be fitted by minimizing the error sum of squares between the measured suction depth and the modeled suction depth using an example technique (e.g., as an example, Broyden-Fletcher-Goldfarb-Shanno algorithm). Equation (B) includes parameters k0, k1, τ, η0, and η1 of the example elastic model 252. The best parameters can be obtained by fitting the suction depth data points. In equation (B), F0 can represent the force applied to the oocyte 110. For example, F0 can represent the force associated with a specific suction depth (hereinafter referred to as "depth"). In the example, different suction depths associated with different times (e.g., images) can be used together with the associated individual forces to determine the parameters. Parameters k0 and k1 simulate the solid-like behavior of the oocyte 110, while η0 and η1 simulate the liquid-like behavior of the oocyte 110. Therefore, the example elastic model 252 can take into account two behaviors associated with the oocyte 110.
[0110]
[0111] in
[0112] The parameters k0 and k1 can describe the "instantaneous elongation" that the oocyte 110 experiences when a force is applied to the oocyte 110. The instantaneous elongation corresponds to or is proportional to 1 / (k0+k1) and can be considered a measure of the "relaxation" of the elastic component of the oocyte 110, or the amount of force that can be applied to the oocyte 110 before significant resistance is exhibited. The parameter k1 can be considered a general measure of stiffness and can represent the tightness of the connection of proteins in the cytoplasm or zona pellucida. The parameter η1 can be considered a measure of the continued deformation of the zona pellucida in response to the applied force calculated according to equation (A). As with the linear elastic solid model, η1 is responsible for maintaining the shape change of the elongated oocyte 110 at the molecular level after the spring component is fully extended. The parameter τ represents the speed at which the oocyte 110 deforms (e.g., enters the pressure tool 112) after the initial instantaneous elongation. η0 can be considered a measure of the viscosity of the fluid in the space (e.g., PVS) between the zona pellucida and the inner cell of the cytoplasm or oocyte 110.
[0113] In addition to using negative pressure, other non-limiting examples of causing oocyte 110 to move or deform can be used and fall within the scope disclosed herein. For example, positive pressure on oocyte 110 can be used to expel oocyte 110 or otherwise deform, hold or disturb different parts of oocyte 110. Example parts can include zona pellucida, cytoplasm, or parts of oocyte 110 near its surface. Other forms of force (e.g., optical pressure) can also be applied to oocyte 110.
[0114] In some examples, the pressure or force applied to the oocyte 110 is appropriately adjusted by the pressure tool 112 to avoid adverse effects on the oocyte 110. For example, too high a pressure applied to the oocyte 110 may damage the structure of the oocyte 110 and reduce the viability of the oocyte 110. In some embodiments, the pressure applied to the oocyte 110 is between -0.01psi and -0.5psi (0.01psi to 0.5psi if positive pressure is applied). In some aspects, the pressure applied to the oocyte 110 by the pressure tool 112 is adjusted according to how many days (e.g., 1, 2, or 3 days) have passed since fertilization. In some embodiments, the inner diameter of the pressure tool 112 is between 40μm and 70μm and the applied pressure can be adjusted based on the inner diameter of the pressure tool 112 to generate an appropriate level of force to be applied to the oocyte 110.
[0115] Sample User Interface
[0116] refer to Figures 3A-3C , will describe Figure 1 Illustrative applications of the oocyte analysis system 100.
[0117] Figure 3AAn example user interface 300A is illustrated, which includes Figure 1 The processing results of the example oocyte analysis system 100 are shown in FIG. Figure 3A As shown, the oocyte quality information 132 generated by the oocyte analysis system 100 can be presented via a user interface 300A. The user interface 300A can be presented as a web page accessible by a browser or as a user interface of an application.
[0118] As shown, user interface 300A receives oocyte quality information 132 from system 100 described herein. Information 132 can display the corresponding quality indicators of a large number of oocytes extracted from the patient. In the illustrated example, user interface 300A presents summary information based on the analysis of 5 oocytes. User interface 300A indicates that "oocyte 3" has the highest possibility of blastocyst formation. The indication can be based on the possibility of blastocyst formation determined by a machine learning model (e.g., model 130).
[0119] User interface 300A can react and check the user input that is associated with detailed analysis.In certain embodiments, detailed analysis may include the subset of all features or features mentioned above.Detailed analysis may include the graphical representation associated with oocyte 3, for example, from the image of image sequence.In certain embodiments, image sequence may be presented as movie or animation in user interface 300A.
[0120] In some examples, the oocyte quality information 132 may be linked or compared to an assessment result obtained from preimplantation genetic testing (PGT) or implantation of the same oocyte sample. For example, for oocytes that the oocyte quality information 132 indicates are capable of forming blastocysts (e.g., "good" oocytes), preimplantation chromosome screening (PGT-A) may be performed to derive the euploidy rate of oocytes selected by the oocyte analysis system 100. As another example, for oocytes that the oocyte quality information 132 indicates are unable to form blastocysts (e.g., "bad" oocytes), preimplantation chromosome screening (PGT-A) may also be performed to derive the aneuploidy rate of oocytes selected by the oocyte analysis system 100.
[0121] As another example, oocytes that are indicated by oocyte quality information 132 as being capable of forming blastocysts (e.g., "good" oocytes, such as having a probability greater than a threshold) may be further evaluated after implantation. For example, this can be used to measure the predictive ability of the oocyte analysis system 100 for oocyte survival. The probability of embryo implantation can be obtained based on the "good" oocytes indicated by the oocyte analysis system 100 to evaluate the predictive power of the oocyte analysis system 100 for embryo implantation. Advantageously, by connecting different stages of embryo quality assessment, the oocyte analysis system 100 can be improved based on the assessment results from PGT-A and implantation. For example, a lower probability of embryo implantation may indicate the need to train the machine learning model 130 by adjusting the parameters of the machine learning model 130 using different subsets of the features 122.
[0122] Figure 3B An example user interface 300B is illustrated that allows Figure 1 The example oocyte analysis system 100 receives user input, such as patient information. Figure 3B As shown, user interface 300B can allow the user to add, edit and store patient information in oocyte analysis system 100.As shown, the patient information displayed on user interface 300B is assigned patient ID (e.g., 1234567890).Patient ID can be used to associate a specific oocyte with a specific patient.Patient information is displayed under patient ID, such as the creation date of height, weight and patient information.Extra patient information can be added by pressing the "+ Add" button on the upper part of user interface 300B.
[0123] although Figure 3B 300B can be used to analyze the quality of one or more specific oocytes. In some embodiments, the oocyte analysis system 10 ...
[0124] The user interface 300B may also receive user input that causes the exemplary oocyte analysis system 100 to analyze a particular oocyte to generate oocyte quality information 132 about the particular oocyte. Additionally, the user interface 300B may prompt the user to check or recalibrate the position of a camera used to capture the sequence of images 102 to analyze the quality of one or more oocytes. In some embodiments, the oocyte quality information 132 generated by the exemplary oocyte analysis system 100 may be presented to the user as described below.
[0125] Figure 3C An example user interface 300C is illustrated, which presents Figure 1 The processing results of the example oocyte analysis system 100 are shown in FIG. Figure 3C As shown, the user interface 300C may present different oocyte quality information 132 based on the analysis performed by the oocyte analysis system 100. Figure 3C As shown, the oocyte quality information 132 is presented as a score indicating the quality of the oocyte. In addition, a textual description summarizing the quality can be presented. For example, the textual description can be based on a quality indicator and can be selected from pre-stored textual expressions or terms (e.g., normal, good, etc.) or determined using a language model (e.g., a large language model).
[0126] The left side shows the analysis results of oocytes with good scores (such as 95), which may represent that the oocyte is very likely to form a "usable" blastocyst. In contrast, the right side shows the analysis results of oocytes with poor scores (such as 14), which may represent that the oocyte is likely not to form a "usable" blastocyst. In the middle, user interface 300C shows the analysis results of oocytes with "normal" scores (such as 60), which may represent that the oocyte is more likely to form a usable blastocyst.
[0127] After receiving the analysis result of a specific oocyte, the user interface 300C can allow the user to view the analysis results of other oocytes or prompt the example oocyte analysis system 100 to analyze the quality of oocytes that have not been analyzed. Specifically, the user can view the analysis results of another oocyte by clicking the "Next Oocyte" button, or view the analysis results of the previously analyzed oocyte by clicking the "Back" button.
[0128] Example Flowchart
[0129] Figure 4 4 is a flow chart of an example process 400 for determining an indicator of oocyte quality. All or at least some portions of process 400 may be performed, for example, by Figure 1The process 400 can be implemented by the oocyte analysis system 100 of the present invention. As described above, the process 400 can be used to determine the viability of the oocyte 110 (e.g., whether it will form a blastocyst) without resorting to the manual and subjective evaluation of the embryologist. Therefore, the process 400 can be used to achieve more objective, time-saving and automated oocyte 110 quality collection.
[0130] At block 402, the system acquires images forming a sequence of images of an oocyte. As described above, the images may be captured by a microscope camera and depict a series of events showing deformation or movement of the oocyte caused by a tool (eg, a micropipette) applying force to the oocyte.
[0131] At frame 404, the system obtains the pressure value associated with the oocyte being sucked into the tool. For example, the pressure value may include the pressure or force applied to the oocyte during the process that the oocyte 110 is sucked into the pressure tool. In certain embodiments, the pressure value may remain the same during the whole suction process. In certain embodiments, the pressure value may change in some way (for example, lower pressure applies higher pressure subsequently).
[0132] At block 406, the system determines morphological features associated with the oocyte based on the images obtained to form the image sequence. Illustratively, the morphological features associated with the oocyte include the aspiration depth of the oocyte, the size and / or length of the cytoplasm, the size and / or length of the zona pellucida, and the diameter length of the tool used to aspirate the oocyte.
[0133] As mentioned above Figure 2A As discussed in, morphological features can be obtained using computer vision techniques and or machine learning techniques. In certain embodiments, one or more neural networks can be trained to output image segmentation shields specific to certain morphological features. In this way, the size or length associated with a part of an oocyte can be identified. In certain embodiments, image segmentation can include two stages, and two-dimensional (2D) modeling is followed by three-dimensional (3D) modeling. In the two-dimensional modeling stage, features associated with two-dimensional images are extracted. In the 3D modeling stage, features extracted from 2D images can be concatenated to form time series data, wherein each time series data includes features extracted from a 2D image in a 2D image sequence. A machine learning model (e.g., a deep learning model) can classify the features generated by the 3D modeling stage, such as in box 410.
[0134] At block 408, the system determines mechanical characteristics associated with the oocyte. For example, the system determines parameters that represent deformation or movement of the oocyte 110 during the image sequence. In this example, the parameters may be related to Figure 2B This is related to the elastic model described in .
[0135] In box 410, the system uses a machine learning model to determine an index representing oocyte quality. The system provides features as inputs to the machine learning model, such as a series of features determined for an image or a sequence of features determined for each image. In some embodiments, the machine learning model 130 can be a support vector machine. In some embodiments, the model can be a deep learning model (e.g., a neural network). In some embodiments, a subset of features can be provided. For example, one feature, two features, three features, 10 features, etc. can be provided.
[0136] The indicator may include at least information indicating oocyte blastocyst formation. In addition, one or more indicators may indicate aneuploidy and / or implantation associated with the oocyte.
[0137] In some embodiments, one or more indicators may indicate whether the oocyte 110 will form a "good" blastocyst, where a "good" blastocyst may represent an associated Gardner embryo / blastocyst grade greater than 3CC. Additionally and optionally, in addition to using the example morphological features and the example mechanical features to determine indicators representing oocyte quality, the machine learning model may also use the patient's clinical information to determine oocyte quality information. As described in Figure 1 As discussed in the related discussion, clinical information may include age, BMI of the patient, and / or other clinical information, such as CP and MII associated with the oocyte (eg, developmental stage associated with the oocyte).
[0138] After determining the indicator indicative of oocyte quality, process 400 may return to block 402 to determine quality information for another oocyte.
[0139] Figure 5 is a flow chart of an example process 500 for selecting a subset of a plurality of oocytes based on an indicator indicative of quality. All or at least some portions of process 500 may be performed, for example, by Figure 1 The oocyte analysis system 100 of the present invention is implemented. The process 500 can be provided for determining the viability between multiple oocytes without resorting to the time-consuming and subjective assessment of embryologists. Therefore, the process 500 can be used to realize the more objective, time-saving and automated collection of a large number of oocyte qualities.
[0140] At block 502, the system obtains a sequence of images and pressure values associated with a plurality of oocytes. As described herein, the sequence of images may depict oocytes deformed due to application or pressure via a pressure tool.
[0141] At block 504, the system obtains an indicator representing the quality of a single oocyte in the plurality of oocytes. Determining the indicator is described in more detail above and may represent the likelihood of each oocyte forming a blastocyst.
[0142] At box 506, the system selects a subset of oocytes from the plurality of oocytes. For example, the system may identify a maximum threshold number of oocytes based on the likelihood that each oocyte will form a blastocyst. As another example, the system may aggregate or otherwise combine the indicators of each oocyte. For this example, the system may select a maximum threshold number of oocytes based on the aggregated or combined indicators. As described above, the indicators may represent successful outcomes of blastocyst formation and later stages or indicators associated with chromosomal abnormalities (e.g., PGT-A, euploidy), etc.
[0143] At block 508, the system outputs and / or presents information associated with the selected oocyte subset. The information may be presented via a graphical user interface (GUI), such as Figure 3A The user interface 300A shown in FIG. Figure 3A and Figure 3C As shown, the information can indicate whether a specific oocyte is likely to form a blastocyst. In addition, other information related to a selected oocyte subset can be presented to the user. For example, morphological features, mechanical features and / or clinical information of the patient from which the oocyte is obtained can be presented. By using the oocyte analysis system 100 implementation process 500, viability information related to a large number of oocytes can be obtained in a time-saving manner.
[0144] Example System
[0145] Figure 6 The general architecture of an example system is depicted. In some embodiments, the system can be used to perform the functions described herein. In some embodiments, the system can be an oocyte analysis system 100, which includes an arrangement of computer hardware and software configured to implement features of the present disclosure. The oocyte analysis system 100 can include: Figure 6 However, it is not necessary to show all of these components to provide implementation of the present disclosure.
[0146] As shown, the oocyte analysis system 100 includes a processor 602, a pressure tool 604 (e.g., Figure 1 pressure tool 112), a network interface 606, an image sensor 608 (e.g., one or more for capturing Figure 1 The image sequence 102 of the microscope camera) and data storage 610 can all communicate with each other via a communication bus 612. In some embodiments, the pressure tool 112 may not be included and the system 100 may represent a back-end processing system. The network interface 606 can provide a connection to one or more networks or computing systems, thereby enabling the oocyte analysis system 100 to receive data from other computing systems, interfaces (e.g., Figure 3AThe user interface 300A) or service receives or transmits information and instructions. In some embodiments, the oocyte analysis system 100 can be configured to process requests from other devices or modules, such as requests to analyze oocyte quality. The data store 610 can illustratively be any non-transitory computer-readable data store, and in various embodiments can store Figure 6 Any or all elements depicted in are loaded into memory 614.
[0147] The processor 602 may also communicate with the memory 614. The memory 614 may contain computer program instructions (grouped into modules or components in some embodiments) that the processor 602 may execute to implement one or more embodiments. The memory 614 typically includes RAM, ROM, and / or other permanent, auxiliary, or non-transitory computer-readable media. The memory 614 may store an operating system 616, which provides computer program instructions for use by the processor 602 in the general management and operation of the oocyte analysis system 100. The memory 614 may also store specific computer executable instructions and other information (which may be referred to as "modules" or "engines" herein) to implement the features of the present disclosure. For example, the memory 614 may include a feature pre-processing engine 632 and a machine learning model 634, which may implement the features of the present disclosure as described above. The memory 614 may also store, for example, a user interface module 618, which may implement the Figure 3A In addition, the memory 614 can store a library 620 (e.g., for storing parameters of different types of machine learning models) and features 630 that can be extracted by a feature preprocessing engine 632. When performing various operations, all modules or components loaded into the memory 614 can also be stored in the data storage 610.
[0148] will be recognized, Figure 6 Many of the components described in the present disclosure are optional, and embodiments of the oocyte analysis system 100 may or may not combine components. Furthermore, the components need not be distinct or discrete. The components may also be reorganized. In some embodiments, the components described as part of the oocyte analysis system 100 may additionally or alternatively be included in other computing devices, such that some features of the present disclosure may be performed by the oocyte analysis system 100, while other features are performed by another computing device.
[0149] All methods and tasks described herein can be performed by a computer system and fully automated. In some cases, a computer system may include multiple different computers or computing devices (e.g., physical servers, workstations, storage arrays, cloud computing resources, etc.) that communicate and interoperate over a network to perform the functions described. Each such computing device typically includes a processor (or multiple processors) that executes program instructions or modules stored in a memory or other non-transient computer-readable storage medium or media or device (e.g., solid-state storage device, disk drive, etc.). The various functions disclosed herein can be embodied in such program instructions, or can be implemented in a dedicated circuit (e.g., ASIC or FPGA) of a computer system. In the case where a computer system includes multiple computing devices, these devices can be but need not be located in the same place. The results of the disclosed methods and tasks can be persistently stored by converting a physical storage device such as a solid-state storage chip or disk into a different state. In some embodiments, the computer system can be a cloud-based computing system whose processing resources are shared by multiple different business entities or other users.
[0150] The processes described herein or shown in the drawings of the present disclosure may be initiated in response to an event, such as on a predetermined or dynamically determined schedule, on demand when initiated by a user or system administrator, or in response to some other event. When such a process is initiated, a set of executable program instructions stored on one or more non-transitory computer-readable media (e.g., hard disk, flash memory, removable media, etc.) may be loaded into the memory (e.g., RAM) of a server or other computing device. The executable instructions may then be executed by a hardware-based computer processor of the computing device. In some embodiments, such a process or portions thereof may be implemented serially or in parallel on multiple computing devices and / or multiple processors.
[0151] Depending on the embodiment, the specific actions, events, or functions of any process or algorithm described herein may be performed in a different order, may be added, combined, or omitted entirely (e.g., not all described operations or events are necessary for algorithm practice). In addition, in certain embodiments, operations or events may be performed simultaneously, for example, through multithreading, interrupt processing, or multiple processors or processor cores or on other parallel architectures, rather than sequentially.
[0152] The various illustrative logic blocks, modules, routines, and algorithm steps described in conjunction with the embodiments disclosed herein may be implemented as electronic hardware (e.g., ASIC or FPGA devices), computer software running on computer hardware, or a combination of the two. In addition, the various illustrative logic blocks and modules described in conjunction with the embodiments disclosed herein may be implemented or executed by a machine, such as a processor device, a digital signal processor ("DSP"), an application-specific integrated circuit ("ASIC"), a field programmable logic gate array ("FPGA"), or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof, intended to perform the functions described herein. The processor device may be a microprocessor, but in an alternative, the processor device may be a controller, a microcontroller, or a state machine, a combination thereof, or the like. The processor device may include a circuit configured to process computer executable instructions. In another embodiment, the processor device includes an FPGA or other programmable device that performs logic operations without processing computer executable instructions. The processor device may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in combination with a DSP core, or any other such configuration. Although this article is primarily described with respect to digital technology, the processor device may also primarily include an analog component. For example, some or all of the rendering techniques described herein may be implemented in analog circuitry or mixed analog and digital circuitry. The computing environment may include any type of computer system, including but not limited to a microprocessor-based computer system, a mainframe computer, a digital signal processor, a portable computing device, a device controller, or a computing engine within a device, to name a few.
[0153] The elements of the methods, processes, routines or algorithms described in conjunction with the embodiments disclosed herein may be directly embodied in hardware, in software modules executed by a processor device, or in a combination of the two. The software module may be stored in a RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, register, hard disk, removable disk, CD-ROM or any other form of non-transient computer-readable storage medium. An exemplary storage medium may be coupled to a processor device so that the processor device can read information from the storage medium and write information to the storage medium. Alternatively, the storage medium may be integrated into the processor device. The processor device and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. Alternatively, the processor device and the storage medium may be in a user terminal as discrete components.
[0154] Conditional language used herein, such as "can," "might," "may," "for example," and the like, unless specifically stated otherwise, is generally intended to convey that certain embodiments are present while other embodiments do not include certain features, elements, or steps. Thus, such conditional language is generally not intended to imply that features, elements, or steps are required for one or more embodiments, or that one or more embodiments must include logic for deciding whether, with or without additional inputs or cues, an element or step is included or will be performed in any particular embodiment. The terms "comprise," "include," "have," and the like are synonymous and are inclusive in an open-ended manner and do not exclude other elements, features, behaviors, operations, and the like. Similarly, the word "or" is used in an inclusive sense (rather than an exclusive sense) so that, for example, when using a list of elements, the word "or" means one, some, or all of the elements in the list.
[0155] Disjunctive language, such as "at least one of X, Y, or Z," is understood in some way, unless expressly stated otherwise, in the general context to indicate that an item, term, etc., is either Y or Z, or any combination thereof (e.g., X, Y, or Z). Thus, such disjunctive language is generally not intended and should not imply that certain embodiments require at least one of X, at least one of Y, and at least one of Z to be present.
[0156] Although the above detailed description has shown, illustrated and pointed out the novel features applied to various embodiments, it is understood that various omissions, substitutions and changes in form and details may be made to the described devices or algorithms without departing from the spirit of the present disclosure. It is recognized that some embodiments described herein may be embodied in a form that does not provide all of the features and benefits set forth herein, because some features can be used or practiced separately from other features. All changes that fall within the meaning and range of equivalency of the scope of the claims should be included within their scope.
Claims
1. A method implemented by a system of one or more computers, characterized in that include: acquiring a plurality of images forming an image sequence associated with a time period, the image sequence depicting an oocyte and a portion of a tool applying pressure to the oocyte, wherein each image is associated with a respective value of pressure applied to the oocyte at a respective time of image capture; Determining features associated with the oocyte based on the image and the pressure value, the features including morphological features representing measurements of the oocyte over a period of time and mechanical features representing deformation of the oocyte over a period of time; as well as determining, via a machine learning model based on input values comprising at least a subset of features, one or more indicators indicative of the quality of the oocyte, wherein a particular indicator is indicative of blastocyst formation, One or more of the pointers are configured to be presented via an interactive user interface.
2. The method according to claim 1, wherein: The oocytes depicted in the images are mammalian.
3. The method according to claim 1, wherein: The image depicts the oocyte being sucked into the partial tool.
4. The method according to claim 1, wherein: The mechanical characteristic represents an elasticity of the oocyte, and wherein the mechanical characteristic is determined based on a depth of suction of the oocyte into a portion of a tool.
5. The method according to claim 4, wherein: The mechanical characteristics represent the viscoelastic behavior of the oocyte.
6. The method according to claim 1, wherein: The morphological feature at least represents the suction depth of the oocyte into the tool part in the corresponding image.
7. The method according to claim 1, wherein: One of the morphological features is derived by normalizing the depth of aspiration of the oocyte into the tool portion of the corresponding image by the inner diameter of the tool.
8. The method according to claim 1, wherein: The morphological characteristics represent a first measurement associated with the cytoplasm of the oocyte and a second measurement associated with the zona pellucida of the oocyte.
9. The method according to claim 1, wherein: The tool portion is a micropipette, and wherein the micropipette is adjacent to the oocyte during a period of time and is configured to apply pressure.
10. The method of claim 1, wherein: A first image associated with the beginning of a time period depicts the oocyte prior to application of pressure, and wherein a second image following the first image in the sequence of images depicts the oocyte with pressure applied.
11. The method of claim 1, wherein: The pressure value is selected from a range of pressures applied over a period of time.
12. The method of claim 1, wherein: The machine learning model is a support vector machine.
13. The method of claim 1, wherein: The machine learning model is a neural network.
14. The method of claim 1, wherein: Also includes: calculating a measurement of the pressure exerted on the oocyte based on the pressure value and geometric information associated with the tool portion; as well as The mechanical characteristics are obtained by fitting an aspiration depth curve using a linear elastic method, wherein the fitted aspiration depth curve is based on the measured values of the aspiration depth and pressure of the oocyte into the pressure tool based on the corresponding images.
15. The method of claim 1, wherein: The input values further include clinical information of a person associated with the oocyte, the clinical information including one or more of age, body mass index, developmental stage associated with the oocyte, and cryopreservation information.
16. An oocyte quality analysis system, characterized in that: A non-transitory computer storage medium comprising one or more processors and storing instructions that, when executed by the one or more processors, cause the one or more processors to: acquiring a plurality of images forming an image sequence associated with a time period, the image sequence depicting an oocyte and a portion of a tool applying pressure to the oocyte, wherein each image is associated with a respective value of pressure applied to the oocyte at a respective time of image capture; Determining features associated with the oocyte based on the image and the pressure value, the features including morphological features representing measurements of the oocyte over a period of time and mechanical features representing deformation of the oocyte over a period of time; as well as determining, via a machine learning model based on input values comprising at least a subset of features, one or more indicators indicative of the quality of the oocyte, wherein a particular indicator is indicative of blastocyst formation, One or more of the indicators are configured to be presented via an interactive user interface.
17. The oocyte quality analysis system according to claim 16, characterized in that: The image depicts the oocyte being sucked into the partial tool.
18. The oocyte quality analysis system according to claim 16, characterized in that: The mechanical characteristic represents an elasticity of the oocyte, and wherein the mechanical characteristic is determined based on a depth of suction of the oocyte into a portion of a tool.
19. The oocyte quality analysis system according to claim 18, characterized in that: The mechanical characteristics represent the viscoelastic behavior of the oocyte.
20. The oocyte quality analysis system according to claim 16, characterized in that: The morphological feature at least represents the depth of aspiration of the oocyte into the tool portion of the corresponding image.
21. The oocyte quality analysis system according to claim 16, characterized in that: The morphological characteristics represent a first measurement associated with the cytoplasm of the oocyte and a second measurement associated with the zona pellucida of the oocyte.
22. The oocyte quality analysis system according to claim 16, characterized in that: The pressure value is selected from a range of pressures applied over a period of time.
23. The oocyte quality analysis system according to claim 16, characterized in that: The instructions further cause one or more of the processors to: calculating a measure of the pressure exerted on the oocyte based on the pressure value and geometric information associated with the tool portion; and The mechanical characteristics are obtained by fitting an aspiration depth curve using a linear elastic method, wherein the fitted aspiration depth curve is based on the measured values of the aspiration depth and pressure of the oocyte into the pressure tool based on the corresponding images.
24. The oocyte quality analysis system according to claim 16, characterized in that: The input values further include clinical information of a person associated with the oocyte, the clinical information including one or more of age, body mass index, developmental stage associated with the oocyte, and cryopreservation information.
25. A method implemented by a system of one or more computers, characterized in that: include: acquiring a plurality of image sequences depicting respective oocytes, wherein each of said image sequences depicts each of said oocytes and a portion of a tool applying pressure to said oocyte, and wherein pressure values associated with images included in said image sequences are acquired; Determining one or more indicators representing oocyte quality of each of the oocytes by a machine learning model, wherein particular indicators represent blastocyst formation, and wherein input values of the machine learning model for each of the oocytes include morphological features representing measurements of the oocyte and mechanical features representing deformation of the oocyte during application of pressure to the oocyte; and The selection is based on the indicator, and information indicating the selection of one or more of the oocytes is output.
26. The method implemented by a system of one or more computers according to claim 25, characterized in that: The image depicts the oocyte being sucked into the partial tool.
27. The method implemented by a system of one or more computers according to claim 25, characterized in that: The mechanical characteristic represents an elasticity of the oocyte, and wherein the mechanical characteristic is determined based on a depth of suction of the oocyte into a portion of a tool.
28. The method implemented by a system of one or more computers according to claim 25, characterized in that: The morphological characteristics represent a first measurement associated with the cytoplasm of the oocyte and a second measurement associated with the zona pellucida of the oocyte.
29. The method implemented by a system of one or more computers according to claim 25, characterized in that: Information representing the selection is output via an interactive user interface, and wherein the interactive user interface: presenting a graphic depiction of one or more of said oocytes; selecting a particular oocyte from among the one or more oocytes in response to user input; and Detailed information related to the particular oocyte is presented.
30. The method implemented by a system of one or more computers according to claim 25, characterized in that: Information representative of the selection is output via an interactive user interface, and wherein the interactive user interface presents a respective text description of each of the oocytes summarizing one or more of the oocytes.
Citation Information
Cited By
Method implemented by computer system, and quality analysis system and computer-readable medium
WO2026001068A1
Computer system-implemented method and oocyte quality analysis system
WO2026199764A1
