Bio-printing method and system

Abstract

In a bio-printing system and method, a bio-printer may include one or more reservoirs, a printhead, and a printing surface onto which bio-prints are dispensed. An optical system may collect imaging data of a plurality of cells within a crosslinkable material before, during, and / or after the plurality of cells pass through the bioprinting system. A computing system may process the imaging data to determine a number, distribution, concentration, and / or morphology of the plurality of cells. The machine learning system compares characteristics of the plurality of cells before, during and / or after the plurality of cells pass through the bioprinting system to a user established cellular morphology corresponding to physical characteristics of cells and / or cell aggregates within a predetermined tolerance. Additionally or alternatively, the machine learning system compares characteristics of a plurality of cells at one stage in the bioprinting system to characteristics of a plurality of cells at another stage in the bioprinting system.

Description

[0001] Cross-reference to related applications This application claims the benefit of U.S. Provisional Application No. 63 / 449,291, entitled “BIOPRINTING METHODS AND SYSTEMS”, filed March 1, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This invention relates to methods and apparatus for quality control of a three-dimensional (3D) bioprinting system. More specifically, it relates to methods and apparatus for monitoring different stages of a 3D bioprinting system. Even more specifically, it relates to computer vision and machine learning systems that provide real-time visual images of multiple cells at different locations within a bioprinting platform to assess various characteristics of the multiple cells, including but not limited to the number, distribution, concentration, quality, and / or morphology of the multiple cells. Background Technology

[0003] The 3D bioprinting process has multiple stages, including the so-called pre-printing stage, printing stage, and post-printing stage. It is desirable to provide a system capable of monitoring the different stages of the 3D bioprinting process to assess how the operation at each stage affects print quality. Such a system could enable improvements in the operation at each stage, thereby facilitating the production of suitable 3D bioprinted tissues for therapeutic purposes. Summary of the Invention

[0004] This invention addresses the aforementioned problems in the art by integrating computer vision and deep learning into a three-dimensional (3D) bioprinting platform, thereby enabling direct monitoring of the operation and flow of various materials at different stages within the bioprinting system. In embodiments, materials may include, for example, cell-loaded hydrogels and other crosslinkable materials. In embodiments, different stages of the bioprinting system may include one or more reservoirs, a printhead, and a print surface.

[0005] Embodiments of the present invention enable non-contact sensing of the properties of multiple cells flowing through the bioprinting system before, during, and / or after printing. Embodiments also achieve automated analysis, which in turn enables the derivation of quantitative measurements of the properties of the multiple cells. These measurements can be used as quality control and / or quality assurance parameters. In some embodiments, such quantitative measurements include real-time, high-level quantitative analysis of the biological material (e.g., cells) before, during, and / or after printing to qualitatively assess cell quality in terms of the consistency of the biological material. In some embodiments, such quantitative measurements include real-time, high-level quantitative analysis of other objects within the material flow.

[0006] In embodiments, the material stream comprises at least one cross-linkable material, and preferably at least one hydrogel, optionally further comprising at least one biomaterial, such as a plurality of cells in a biocompatible material. In embodiments, the plurality of cells is selected from the group comprising or consisting of a single cell suspension, a cell aggregate, a cell cluster, a cell conglomerate, a cell spheroid, a cell organoid, or a combination thereof. In embodiments, the material stream comprises microparticles.

[0007] In embodiments, the plurality of cells comprises cells from an endocrine and exocrine gland selected from the group consisting of pancreas, liver, thyroid, parathyroid, pineal gland, pituitary gland, thymus, adrenal gland, ovary, testis, an enteroendocrine cell, a stem cell, a stem cell-derived cell, or a cell engineered to secrete a bioactive agent of interest. In embodiments, the plurality of cells releases cell-derived extracellular vesicles in the form of exosomes comprising a therapeutic protein or nucleic acid. In embodiments, the biocompatible material is selected from the group consisting of alginate, collagen, decellularized extracellular matrix, hyaluronic acid, PEG, fibrin, gelatin, GEL-MA, silk worm, chitosan, cellulose, PCL, PLA, POEGMA, and combinations thereof.

[0008] In embodiments, the print head can comprise one or more transparent channels, and the optical system can monitor the material stream through at least one of the one or more transparent channels, preferably wherein the print head comprises a transparent nozzle or dispensing channel. In embodiments, the print head can be an extrusion print head or a co-extrusion print head, including but not limited to multi-material extrusion, multi-axis extrusion or melt extrusion, an inkjet print head, an electro-writing print head, or a microfluidic print head, including but not limited to on-chip cross-linking, microfluidic stereolithography, or microfluidic spinning.

[0009] In general, the print head will comprise one or more inputs fluidically connected to one or more reservoirs, one or more dispensing channels for depositing the material stream onto a print surface, and preferably one or more transparent channels or regions. The cross-linkable material within the material stream can be cross-linked within the print head, upon dispensing from the print head, and / or upon deposition on the print surface, as is known in the art. The print surface can further comprise, for example, a cross-linker bath.

[0010] In embodiments, the optical system can comprise one or more microscopes, one or more still cameras, and / or one or more video cameras. There can be an optical system for each stage of the bioprinting system, from the reservoir to the print head to the print surface.

[0011] In exemplary embodiments, the print head can comprise a plurality of transparent channels, and the optical system can comprise an equal number of microscopes, still cameras, and / or video cameras.

[0012] In embodiments, as the plurality of cells traverse the bioprinting system, the machine learning system identifies one or more deviations of the flow of the plurality of cells from the user-established material flow parameters at each of the plurality of stages. In embodiments, and in response to the identified one or more deviations, the machine learning system identifies whether it is necessary to adjust the material flow parameters. In embodiments, the machine learning system can adjust the material flow parameters in response to the cumulative deviations exceeding a predetermined amount. In embodiments, the machine learning system adjusts the material flow parameters so as to maintain suitable physical properties of the plurality of cells within predetermined tolerances.

[0013] In embodiments, the machine learning system performs panoramic segmentation and / or semantic segmentation and / or instance segmentation on the stream of images of the flow of the plurality of cells traversing the bioprinting system. In embodiments, the panoramic segmentation and / or semantic segmentation and / or instance segmentation enables detection of the location of one or more objects within the material flow. In embodiments, the panoramic segmentation and / or semantic segmentation and / or instance segmentation enables visual estimation of the shape and / or size of one or more objects within the material flow. In embodiments, the panoramic segmentation and / or semantic segmentation and / or instance segmentation enables visual estimation of the regular amount and / or distribution of biological material (e.g., the plurality of cells) within the material flow.

[0014] In embodiments, the apparatus comprises a three-dimensional (3D) bioprinting print head. In embodiments, the 3D bioprinting print head comprises a plurality of channels to selectively provide a respective plurality of materials of the material flow.

[0015] In embodiments, the at least one cross-linkable material comprises a hydrogel. In embodiments, the material flow further comprises at least one biological material; preferably wherein the at least one biological material comprises a plurality of cells. In embodiments, the plurality of cells is selected from the group consisting of, or consisting of, a single cell suspension, a cell aggregate, a cell cluster, a cell conglomerate, a cell spheroid, a cell organoid, or a combination thereof. In embodiments, the material flow further comprises microparticles. In embodiments, the material flow further comprises a dye, a pigment, or a colloid. In embodiments, the presence of cells in the material flow acts as a contrast agent to facilitate measurement of physical properties of the bioprinted fiber.

[0016] In another embodiment, the computing system uses the results of the comparison to control the material flow by adjusting the pressure of the material flow within the bioprinting system. In embodiments, the system further comprises a pressure controller that controls the material flow and pressure through the print head during printing of the printed fiber in response to the results of the comparison.

[0017] In embodiments, the machine learning system can include a neural network selected from the group consisting of a convolutional neural network (CNN), a fully convolutional network (FCN), a region-based CNN (R-CNN), Mask R-CNN, a “You Only Look Once” (YOLO) based model, and a Transformer-based instance segmentation model. Further, in embodiments, a flow estimation network including, but not limited to, a flow estimation network in the FlowNet series (e.g., FlowNet, FlowNet2), a recurrent affinity field transform (RAFT), and self-learning multi-frame unsupervised RAFT with full-image warping (SMURF) can be used. Yet further, in embodiments, joint prediction of cell segmentation and flow estimation using a technique called SegFlow can be used.

[0018] In embodiments, determining further includes performing a panoptic segmentation and / or a semantic segmentation and / or an instance segmentation on the streaming images of the material flow using a machine learning system. In embodiments, the panoptic segmentation and / or the semantic segmentation and / or the instance segmentation enables identification of cells, debris, and aggregates within the material flow. In embodiments, the panoptic segmentation and / or the semantic segmentation and / or the instance segmentation enables visual estimation of the shape and / or size of one or more objects within the material flow. In embodiments, the panoptic segmentation and / or the semantic segmentation and / or the instance segmentation enables visual estimation of the amount and / or distribution of biological material (e.g., a plurality of cells) within the material flow.

[0019] In embodiments, obtaining includes obtaining streaming images through one or more transparent channels in a three-dimensional (3D) bioprinting head in the apparatus, the 3D bioprinting head producing bioprinted fibers. In embodiments, monitoring includes monitoring a plurality of channels within the 3D bioprinting head, the plurality of channels selectively providing a respective plurality of materials to the material flow. BRIEF DESCRIPTION OF DRAWINGS

[0020] The accompanying drawings, which are incorporated herein and constitute part of this specification, illustrate embodiments of the disclosure and, together with the description, further serve to explain the principles of the disclosure and to enable a person skilled in the relevant art to make and use the disclosure.

[0021] Figure 1 is a high-level diagram of a bioprinting system according to embodiments; FIG. 2 is a high-level flowchart showing different stages before, during, and after a plurality of cells pass through different stages of a bioprinting system according to embodiments; Figures 3A to 3F depicts different geometric and characteristic representations of cells, debris, and aggregates according to embodiments; Figure 4is a flow chart showing how to distinguish between cells, debris and aggregates according to an embodiment; Figure 5 shows a plurality of microwells for characterizing a plurality of cells prior to printing according to an embodiment; Figures 6A to 6D shows an application of the characterization of cells, debris and aggregates according to an embodiment; Figure 7A shows the detection of aggregates according to an embodiment, whereas Figure 7B shows the detection of both cells and aggregates according to another embodiment; Figure 8A and Figure 8B shows the characterization of cells, debris and aggregates prior to printing according to an embodiment; Figure 9A and Figure 9B shows the characterization of cells, debris and aggregates during printing according to an embodiment; Figures 10A to 10H shows the characterization of cells, debris and aggregates after printing according to an embodiment; Figures 11A to 11C shows the characterization of cells, debris and aggregates according to an embodiment; Figure 12A and Figure 12B shows the characterization of cells, debris and aggregates after printing according to an embodiment; Figure 13A and Figure 13B shows the identification of cells, debris and aggregates according to an embodiment; Figure 14A and Figure 14B demonstrates the detection of cells, debris and aggregates according to an embodiment; Figure 15A and Figure 15B demonstrates the detection of cells, debris and aggregates according to an embodiment; Figure 16A and Figure 16B demonstrates the detection of cells, debris and aggregates according to an embodiment; Figure 17A and Figure 17B demonstrates the detection of cells, debris and aggregates according to an embodiment; Figure 18A and Figure 18B demonstrates the detection of cells, debris and aggregates according to an embodiment; Figure 19A and Figure 19B shows the identification of cells, debris and aggregates after printing according to an embodiment; Figure 20A and Figure 20B shows identification of cells, debris, and aggregates after printing according to an embodiment; Figure 17A , Figure 17B , Figure 20A and Figure 20B is an input image, Figure 21A , Figure 21D , Figure 21G and Figure 21J shows detected cells according to an embodiment, while Figure 21B , Figure 21E , Figure 21H and Figure 21K shows a histogram of cell property distribution.

[0022] Figure 21C shows identification of cells and aggregates before printing according to an example; Figure 21F shows identification of cells and aggregates during printing according to an example; Figure 21I shows identification of cells and aggregates after printing according to an example; Figure 21L shows identification of cells and aggregates that outline beads in printed fibers according to an example; Figures 6A to 6D shows a bead quantification model integrated into a print analysis software. DETAILED DESCRIPTION

[0023] Certain illustrative aspects of the systems, apparatuses, and methods consistent with the present disclosure are described herein with reference to the following description and drawings. These aspects are indicative, however, of but a few of the various ways in which the principles of the present disclosure can be employed and the present disclosure is intended to include all such aspects and their equivalents. Other advantages and novel features of the present disclosure can become apparent from the following detailed description when considered in conjunction with the drawings.

[0024] In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the present application. In other instances, well-known structures, interfaces, and processes have not been shown or described in detail in order not to unnecessarily obscure aspects of the present application. It will be apparent, however, to one of ordinary skill in the art having the benefit of this disclosure, that many of those specific details are not required in order to practice the present application and that the specific details disclosed herein should not be interpreted as limiting the scope of the present application. No part of this specification should be interpreted as admitting that any part of this specification is prior art to the present application. Although certain embodiments of the present disclosure are described, these embodiments are likewise not intended to limit the full scope of the present disclosure.

[0025] Figures 11A to 11CA high-level diagram of a bioprinting system 100 according to embodiments is shown. The bioprinting system 100 includes multiple stages, labeled for ease of reference as reservoirs 110 (which can be one or more reservoirs), a print head 120, and a print surface 130 (which can be any surface on which material from the print head 120 can be deposited). In embodiments, the print head can be an extrusion print head or a co-extrusion print head, including but not limited to multi-material extrusion (e.g., Liu et al. Figures 22A to 22F Figures 23A to 23F 2017 Jan; 29(3):10.1002), multi-axis extrusion or melt extrusion (e.g., Loewner et al., Figures 24A to 24P Figures 25A to 25F 2022 DOI 10.3389 / fbioe.2022.896719), inkjet print head (e.g., Lal Roy, Figure 26 2021 DOI 10.1039 / d1lc00524c), electro-writing print head (e.g., Kade and Dalton, Figures 22A to 22F 2020 DOI 10.1002 / adhm.202001232), or microfluidic print head, including but not limited to on-chip crosslinking, microfluidic stereolithography (e.g., Miri et al. Figure 22B . 2018Jul; 30(27): e1800242), or microfluidic spinning (e.g., Lee et al. Figure 22E 2011 10:877-883). In preferred embodiments, the bioprinting system employs on-chip crosslinking, as disclosed in U.S. Patent No. 11,046,930, PCT / CA2014 / 050556, PCT / CA2018 / 050315, PCT / CA2018 / 050315, PCT / CA2018 / 050315, and PCT / CA2022 / 051855, the disclosures of which are expressly incorporated herein by reference.

[0026] Each stage has a respective optical system associated therewith that provides image data for the cells and cell aggregates of each stage. For example, the reservoir 110 can have an optical system 112 associated therewith that can include a microscope, a still camera, and / or a video camera to acquire images of the plurality of cells in the reservoir at an appropriate angle to characterize the plurality of cells prior to printing. These cells can be individual cells; they can be so-called fragments; or they can be cell aggregates. Similarly, the print head 120 can have an optical system 122 associated therewith that can also include a microscope, a still camera, and / or a video camera to acquire images of the cells in the print head (e.g., the cells in the print head nozzles as described in the above-referenced U.S. provisional application and PCT application) at an appropriate angle to characterize the plurality of cells during printing. These cells similarly can be individual cells; they can be so-called fragments; or they can be cell aggregates. Likewise, the print surface 130 can have an optical system 132 associated therewith that likewise can include a microscope, a still camera, and / or a video camera to acquire images of the cells on the print surface at an appropriate angle to characterize the plurality of cells after printing. These cells likewise can be individual cells; they can be so-called fragments; or they can be cell aggregates.

[0027] In embodiments, the bioprinting system 100 can be controlled by a computing system 150 that can include one or more central processing units (CPUs), volatile memory, non-volatile memory, and non-transitory non-volatile storage devices. The computing system 150 can be connected with one or more of the optical systems 112, 122, and 132 through various connections (direct connection, connection through a local area network (LAN), or connection through the cloud).

[0028] The computing system 150 can include a machine learning system 160. In embodiments, the machine learning system 160 can employ one or more graphics processing units (GPUs) having multiple cores to facilitate: the necessary computations for the machine learning system to perform fast computational analysis of data streams or captured images; and the training of models to provide feedback to control the flow of materials in the bioprinting system. In embodiments, the GPUs of the machine learning system 160 can be part of the computing system 150. In embodiments, the machine learning system 160 can use the memory and storage devices that the computing system 150 has. In other embodiments, the machine learning system 160 can have its own volatile memory, non-volatile memory, and / or non-transitory non-volatile storage devices. In embodiments, the model training can involve testing and validation to facilitate optimization and inference by the trained models.

[0029] In embodiments, the machine learning system 160 can include a neural network selected from the group consisting of a convolutional neural network (CNN), a region-based CNN (R-CNN), a Mask R-CNN, a "You Only Look Once" (YOLO) based model, and a Transformer-based instance segmentation model to perform image segmentation on imaging data received from the optical systems 112, 122, and 132. In embodiments, the machine learning system 160 can interact with the computing system 150 to perform a number of user interaction functions as well as system interaction functions. For user interactions, the computing system 150 can provide appropriate graphical displays and graphical user interfaces for interacting with various other components such as the optical systems, the machine learning system 160, and the rest of the bio-printing system 100.

[0030] In a cloud environment, the machine learning system 160 and its user interface can be deployed as a cloud application, e.g., as a website that is login accessible, through which a user can view and interact with a print video online and see the results of the machine learning analysis. In such embodiments, the computing system 150 and the machine learning system 160 can be hosted in the cloud.

[0031] Depending on the embodiment, the computing system 150 can control the optical systems 112, 122, and 132 through one or more graphical user interfaces on one or more displays. Depending on the embodiment, the optical systems 112, 122, and 132 can have their own control devices, whereby these optical systems can position themselves appropriately with respect to the reservoir 110, the print head 120 and its associated feeders and nozzles, and the print surface 130, respectively, in response to commands from the computing system 150.

[0032] Part of the control performed by the computing system 150 involves monitoring access to the control system in the bio-printing system to adjust the flow of material, whether by controlling the pressure in various print head channels (including, e.g., microfluidic print head channels) discussed below, or by controlling the displacement of material through the channels. In embodiments, the control can involve turning on or off or otherwise adjusting the opening and closing of pneumatic valves on the print head. In embodiments, part of the control performed by the computing system 150 involves panoramic segmentation and / or semantic segmentation and / or instance segmentation. In embodiments, the panoramic segmentation and / or semantic segmentation and / or instance segmentation enables a visual estimate of the regular amount and / or distribution of biological material (e.g., a plurality of cells) within the flow of material.

[0033] The computing system 150 can also enable reading from the respective feeds from one or all of the optical systems to enable recording and / or loading one or both of the feeds, and adjusting viewing parameters such as contrast, hue, brightness, and sharpness, among others.

[0034] Depending on the implementation, the machine learning system 160 can characterize the particles in the bioprinting system 100 in a prescribed manner. Alternatively or additionally, the machine learning system 160 can help focus on images from the optical systems 112, 122, and 132 to help characterize the particles. Alternatively or additionally, the machine learning system 160 can provide output that can be fed back in real time to the 3D bioprinting system to adjust the pressure and / or displacement and subsequent flow of material from the reservoir 110 to the print head 120 to the print surface 130, enabling consistent production of high-quality tissue and minimizing loss of expensive biological material and cell input.

[0035] Figure 22A A very high level view of the processes involved in the operation of the bioprinting system 100 in conjunction with the computing system 150 and the machine learning system 160 is shown, according to an implementation. At 210, cell quantification and characterization can be performed prior to printing, for example in the reservoir 110 of the bioprinting system 100. The processes discussed below can be used to quantify and characterize the cells in each well of the microplate. Figure 22D Cells in a microplate are shown, each well containing an aggregate. The processes discussed below with respect to Figures 23A to 23F The processes discussed below can be used to quantify and characterize the aggregates in each well.

[0036] At 212, the cells can optionally be pre-printed on a plate, for example to facilitate aggregation. Again at this stage, the cells can be quantified and characterized. 212 is not required and can be omitted. At 214, the cells are in the print head, awaiting printing. Here, the cells in the print head can again be characterized using the optical systems, computing systems, and machine learning systems in Adv In an implementation, a comparison can be made between the cells in the print head and the cells in the microplate. At 216, the cells can be deposited into a print device, for example Mater the print surface 130 in the bioprinting system 100. At 218, the device structure is characterized. At 220, cell analysis can be completed prior to implantation of the resulting tissue. Depending on the implementation, the same machine learning system 160 or a different machine learning system in Frontiers in Bioengineering and Biotechnology the machine learning system 160 at a given stage can be compared to the results at other stages, and the results at a given stage can be compared to a reference to enable determination of whether and / or how to further train the machine learning system.

[0037] Lab on a Chipis a high-level flow chart for implementing and using computer vision and machine learning tools to provide automated feedback control. At 230, the imaging devices in the optical systems 112, 122, and 132 are controlled to record video and / or still images of the material passing through various stages of the bioprinting system 100. These images are obtained through various operations of the bioprinting system. For example, to generate a training set of data for a machine learning model, the system can be operated with known materials and parameters to obtain known results. Other data sets can also be generated for testing and evaluating the trained system. In embodiments, existing data sets can be used for training.

[0038] At 235, the collected data can be separated into the aforementioned training set and test / evaluation set. At 240, focusing on the training set, features of interest can be labeled to facilitate focusing on those features during training. At 245, the appropriately labeled training set can be applied to a machine learning model of the type mentioned above to train the model, evolving the model to produce correct results from the training data. In some machine learning models, forward propagation of results can be used. In other models, backpropagation can be used.

[0039] At 250, the training results can be analyzed to identify various features of interest, such as the number, distribution, concentration, mass, and / or morphology of the plurality of cells passing through a particular stage of the bioprinting system 100. In embodiments, the morphology can include at least one characteristic selected from the group consisting of cell count, area, perimeter, ellipsoid volume, roundness, smoothness, compactness, circularity, mean diameter, minimum diameter, maximum diameter, mean intensity, and fiber density. The results can then be classified according to the features of interest, level of accuracy, etc. At 255, the test / verification set can be used to evaluate the performance of the trained network. At 260, if the model is satisfactory, at 270, the completed trained model can be deployed. If the model is not satisfactory, additional training samples can be added at 265, and / or the model can be otherwise adjusted. Control then returns to 240 for further training. In embodiments, the addition of the training set can be informed by the nature of the results obtained with the previous training set.

[0040] Advanced Healthcare Materials Facilitates determining how to characterize the plurality of cells passing through the bioprinting system 100 according to embodiments. In embodiments, the morphology of the plurality of cells can be characterized. Adv Mater An aggregate 300 including one or more cells to be characterized is shown. The aggregate 300 can be approximated as a polygon. The polygon will have a centroid, and will have an area. Nature Materials A circle 310 having about the same area and centroid as the polygon is shown. Figure 2AThe aggregate 300 is shown to have areas 311, 313, 315, and 317 outside the circumference of the circle 310. Figure 5 The circle 310 is also shown to have areas 312, 314, 316, and 318 outside the aggregate 300. This circle will have an area The equation for r is solved to enable determination of the circle diameter.

[0041] Figure 4 An ellipsoid volume 320 is shown to have a long diameter D L and a short diameter D S The ellipsoid volume 320 is drawn to have the same volume as the sphere. Approximately, equals the diameter D of the circle.

[0042] Given the foregoing, a roundness value for the aggregate 300 can be obtained. Roundness is the ratio of the area of the circle 310 to the area of the aggregate 300, as seen in Figure 1 which shows the same areas 311, 313, 315, 317 of the aggregate 300 outside the circumference of the circle 310, and the same areas 312, 314, 316, 318 of the circle 310 outside the aggregate 300. The more rounded the aggregate 300, the closer the roundness will be to 1. The more unrounded (e.g., more elliptical) the shape, the closer the roundness will be to 0.

[0043] Figure 1 is a plot that helps characterize the surface complexity of the aggregate 300. The more convex the aggregate 300, the smoother it is. Figure 1 The aggregate 300 is shown with a convex hull 330 wrapped around it. The convex hull 330 is the smallest convex polygon that contains the aggregate 300. Then, the smoothness (or convex hull tightness) of the aggregate 300 is the ratio of the area of the aggregate 300 to the area of the convex hull 330. For a perfectly convex polygon shape of the aggregate 300, the smoothness will be 1. The more concavities in the aggregate 300, and the larger the concavities, the closer the smoothness will be to 0.

[0044] Figure 2B An alternative method for determining the area and / or diameter of the aggregate 300 is shown. In Figures 3A to 3F a circle 350 is wrapped around the aggregate 300. In embodiments, the circle 350 has the same centroid as the aggregate 350. Several diameters 360, 365, and 370 of the aggregate 300 are shown, with diameter 360 being the longest, and diameter 365 being the shortest. Diameter 370 represents the mean of diameters 360 and 365.

[0045] The following table summarizes Figure 3A the metrics described: Figure 3B Figure 3B Figure 3B Figure 3C Figure 3D Figure 3E Figure 3E Figure 3F Figure 3F Figures 3A to 3F Indicator Description Aggregate Area Area of the interior of the aggregate polygon Aggregate Perimeter Aggregate Exterior Perimeter Circumference Diameter Diameter of a circle with the same area as the polygon Ellipsoid Volume Volume of an ellipsoid with the same cross-sectional area as the aggregate, but with a depth given by the circumference diameter Circularity Proportion of the aggregate area within the radius of a circle at the centroid of the aggregate Smoothnessis a flowchart for characterizing a group of cells among a plurality of cells passing through a bioprinting system 100. At 410, it is determined whether the group of cells has a diameter greater than 40 pm. If so, then at 412 it is determined that the group of cells is an aggregate. If not, then at 420 it is determined whether the group of cells has a diameter greater than 25 pm but less than 40 pm. If so, then at 422 it is determined whether the group of cells has a ratio of a long axis of the group of cells to a short axis of the group of cells greater than or equal to 1.3. If so, then at 424 it is determined that the group of cells is a fragment. If not, then at 430 it is determined whether the ratio of the long axis to the short axis just mentioned is less than or equal to 1.3. If so, then at 432 it is determined whether the group of cells has a diameter less than 30 pm. If so, then at 434 it is determined that the group of cells is a single cell. If not, then at 440 it is determined whether the group of cells has a diameter greater than 30 pm and less than 40 pm. If so, then at 442 it is determined whether the group of cells has a circularity greater than or equal to 0.70. If so, then at 444 it is determined that the group of cells is a fragment. If not, then at 450 it is determined whether the group of cells has a circularity greater than 0.70. If so, then at 452 it is determined that the group of cells is an aggregate. Finally, review Ratio of the aggregate area to the area of the convex hull the top of FIG. 6A, if at 420 it is determined that the group of cells does not have a diameter greater than 25 pm and less than 40 pm, then at 460 it is determined whether the group of cells has a diameter less than 25 pm. If so, then at 462 it is determined that the group of cells is a single cell.

[0046] Compactness shows a plurality of microwells, each having a plurality of cells contained therein. The numbers in the figure show the smoothness and circularity of the plurality of cells in each microwell.

[0047] Ratio of the approximate circumference to the aggregate perimeter shows pre-printed cells. FIG. 6B is a slightly magnified view of a portion of Mean Diameter Diameter of the angle bisector between the minimum / maximum diameters Minimum Diameter includes color coding to characterize the plurality of cells shown. Shortest diameter of the aggregate through its centroid is a set of histograms corresponding to the color coding in Maximum Diameter and Longest diameter of the aggregate through its centroid The first histogram shows that various aggregates cluster around a region of relatively low numbers. The second histogram shows, in a manner similar to the first histogram, Figure 4 The third and fourth histograms show that various aggregates tend to have higher circularity and higher smoothness.

[0048] Figure 4 shows an example of cell detection according to one embodiment. Figure 5 ​An example of both cell detection and aggregate detection during the pre-printing phase is shown, according to another embodiment. Among other things, this comparison shows that the latter embodiment identifies significantly more cells than the former embodiment.

[0049] Figure 6A A plurality of cells determined by a machine learning system, encoded in different colors to represent different sizes of structures, is shown, according to an embodiment. Figure 6C The meaning represented by different colors is shown, and includes a table classifying various structures in Figure 6B and as either cells, debris, or cell aggregates.

[0050] Figure 6C and Figure 6D Cell quantification at the print head during printing is shown. Similar to FIG. 6B, Figure 6C , Figure 6D Classification of the plurality of cells in the print head is shown.

[0051] Figure one Cells after printing are shown, and Figure 7A Cell quantification, according to an embodiment, is shown. Similarly, Figure 7B Cells after printing are shown, and Figure 8A Cell quantification after printing is shown.

[0052] Figure 8B Detection of aggregates and cells in the image after printing is shown. According to an embodiment, the cells and aggregates are located in the shell region of the printing device. Figure 8A Cells after printing are shown, and Figure 9A Cell quantification after printing, according to an embodiment, is shown. Similarly, Figure 9B Cells after printing are shown, and Figure 6C Cell quantification after printing, according to an embodiment, is shown.

[0053] Figure 9B Cells after printing are shown, and Figure 10A Quantification of these cells after printing is shown. Similar to Figure 10B , Figure 10C The sampled cell areas are generally smaller and have a smaller radius, and higher roundness and smoothness, are shown.

[0054] Figure 10D and Figures 10E to 10H Another task that can be trained for a machine learning system to perform, according to an embodiment, is shown. Figure 10E Cell detection after printing is shown, and a boundary around each identified cell is provided. Figure 10FWhich cells are in focus are identified and a border is provided around each cell. From these figures, the person of ordinary skill will recognize that the machine learning system according to embodiments can distinguish between in-focus and out-of-focus images from the optical system and work only on the cells that are in focus in that in-focus image.

[0055] Figure 10G And Figure 10H With Figure 11A And Figure 11B Similar to Figure 6D A border is provided around the detected cells after printing, whereas Figure 11C Illustrates the determination of which of the detected cells are actually in focus. Similar to Figure 12A And Figure 12B Similar to Figure 12A With respect to Figure 12B Fewer borders because Figure 13A Borders are only placed around cells that are in focus.

[0056] Figure 13B Illustrates a picture of the cells after printing. Figure 12A The cells that are in focus after printing are made more clear. The machine learning system according to embodiments can be trained to identify cells, debris, and aggregates and can further discriminate which of the detected cells, debris, and aggregates are in focus and in this case highlight them accordingly by contrast. Similar to Figure 12B And Figure 13A Similar to Figure 13B Illustrates a picture of the cells after printing, whereas Figure 12A The cells that are in focus after printing are made more clear.

[0057] Figure 12B And Figure 13B Describes yet another task that the machine learning system according to embodiments can perform. Figure 13A Illustrates a picture of a plurality of cells, debris, and aggregates in a fiber. The cells, debris, and aggregates are in different degrees of focus. Figure 13B Illustrates a picture of the same plurality of cells, debris, and aggregates, but where more of them are in focus. The machine learning system according to embodiments can not only identify cells, debris, and aggregates, but also sharpen their focus.

[0058] Figure 14A Illustrates the cells after printing before the machine learning system detects them. Figure 14B Illustrates detected cells that are in focus and out of focus. These figures thus demonstrate that the machine learning system according to embodiments can detect cells, whether in focus or not, and highlight them appropriately. This detection facilitates quantification and quality review. Figure 14A And Figure 14B、 Figure 15A and Figure 15B and Figure 16A and Figure 16B have similar effects as Figure 16A and Figure 16B respectively. Figure 17A and Figure 17B show cells in fibers according to embodiments.

[0059] Figure 18A 、 Figure 18B 、 Figure 19A and Figure 19B are examples of input Z-stack images taken at different times. These images show the detection of in-focus cells and aggregates in different Z-stack layers of printed fibers at different times. In embodiments, the model detects in-focus cells and aggregates. When the user changes the focus of the microscope to image different layers, in-focus cells and aggregates in different Z-stack layers can be quantified. Figure 20A 、 Figure 20B 、 Figure 17A and Figure 17B show cell detection in images of Figure 20A 、 Figure 20B 、 Figure 21A and Figure 21D respectively. Figure 21G 、 Figure 21J 、 Figure 21B and Figure 21E show histograms that indicate the distribution of properties (area, radius, roundness, and smoothness) are wider than those shown in Figure 21H and Figure 21K . Thus, these graphs indicate that even in the case of irregular shapes, the machine learning system according to embodiments can be trained to detect and quantify cells, debris, and aggregates.

[0060] Material flow: Aspects of the invention include material flows that can be used to print structures so as to be advantageously used as biological materials. As used herein, “biological material” refers to natural or synthetic substances that can be used to build or replace tissue, e.g., human tissue with or without living cells. In the field of bioprinting, the term “biological material” is often synonymous with the term “bioink”.

[0061] The material stream generally includes at least one cross-linkable material, such as a hydrogel, including but not limited to alginate, chitosan, PEGDA, PEGTA, hyaluronic acid (HA), HAMA, collagen, CollMA, gelatin, methacrylated gelatin (gelMA), agarose, gellan gum, fibrin (fibrinogen), PVA, and the like or any combination thereof, as well as non-hydrogels, including but not limited to PCL, PLGA, PLA, and the like or any combination thereof. In preferred embodiments, the material stream includes at least one hydrogel. Non-limiting examples of hydrogels include alginate, agarose, collagen, fibrinogen, gelatin, chitosan, hyaluronic acid-based gels, or any combination thereof. A variety of synthetic hydrogels are known and can be used in embodiments of the systems and methods provided herein. For example, in some embodiments, one or more hydrogels form at least a portion of the structural foundation of a three-dimensional structure for printing. In some embodiments, the hydrogel has the ability to support the growth and / or proliferation of one or more cell types, which can be dispersed within the hydrogel or added to the hydrogel after it is printed in a three-dimensional configuration.

[0062] In embodiments, the hydrogel can be cross-linked by a chemical cross-linking agent. For example, a hydrogel comprising alginate can be cross-linkable in the presence of a divalent cation such as calcium chloride (CaCl2), a hydrogel comprising chitosan can be cross-linked using a multivalent anion such as sodium tripolyphosphate (STP), a hydrogel comprising fibrinogen can be cross-linkable in the presence of an enzyme such as thrombin, and a hydrogel comprising collagen, gelatin, agarose, or chitosan can be cross-linkable in the presence of heat or an alkaline solution.

[0063] In some embodiments, the hydrogel includes alginate. When contacted with divalent cations, alginate forms a solidified colloidal gel (high water content gel or hydrogel). Any suitable divalent cation can be used to form a solidified hydrogel with input materials including alginate. Of the alginate ion affinity series Cd2+> Ba2+> Cu2+> Ca2+> Ni2+> Co2+> Mn2+, Ca2+is the most characteristic and most commonly used to form alginate gels (Ouwerx, C. et al., Polymer Gels and Networks, 1998, 6(5):393-408). Studies have shown that calcium alginate gels form by cooperative binding of Ca2+ions on adjacent poly G blocks, i.e., the so-called “egg-box” model (ISP Alginates, Section 3: Algin-Manufacture and Structure, in Alginates: Products for Scientific Water Control, 2000, International Specialty Products: San Diego, pp. 4-7). G-rich alginate tends to form thermally stable, strong and brittle Ca gels, while M-rich alginate tends to form gels that are less thermally stable, weaker but more elastic. In some embodiments, the hydrogel includes depolymerized alginate.

[0064] In some embodiments, the hydrogel can be crosslinked using a free radical polymerization reaction to generate covalent bonds between molecules. Free radicals can be generated by exposing a photoinitiator to light (typically ultraviolet light), or by exposing the hydrogel precursor to a chemical source of free radicals, such as ammonium persulfate (APS) or potassium persulfate (KPS) in combination with N,N,N,N-tetramethylethylenediamine (TEMED) as an initiator and catalyst, respectively. Non-limiting examples of photo-crosslinkable hydrogels include: methacrylated hydrogels, such as hyaluronic acid methacrylate (HAMA), gelatin methacrylate (GEL-MA), or poly(ethylene glycol) acrylate (PEG-acrylate)-based hydrogels, which are used in cell biology due to their inertness to cells. Polyethylene glycol diacrylate (PEG-DA) is commonly used as a scaffold in tissue engineering because the polymerization reaction occurs rapidly at room temperature, the energy input required is low, it has a high water content, is elastic, and can be customized to include various biomolecules.

[0065] In embodiments, the material stream includes a non-biodegradable polymer. In examples, the input material can be a synthetic polymer, such as polyvinyl acetate (PVA). In embodiments, the material stream can include hyaluronic acid (HA).

[0066] In embodiments, the material flow includes microparticles, which as used herein refers to insoluble particles ranging from about 0.1 um to about 100 um, which are typically composed of polymers, metals, or other inorganic materials. They can be symmetrical (e.g., spherical, cubic, etc.), but this is not required. Microparticles with an aspect ratio of 2: 1 or greater can be considered microrods or microfibers.

[0067] Additional components: The material flow according to embodiments of the present application can include any of a variety of natural or synthetic polymers that support the viability of living cells, including, for example, alginate, laminin, fibrin, hyaluronic acid, polyethylene glycol-based gels, gelatin, chitosan, agarose, or combinations thereof. In embodiments, the subject compositions are physiologically compatible, i.e., conducive to cell growth, differentiation, and communication. In certain embodiments, the input material includes one or more physiologic matrix materials or combinations thereof. By "physiologic matrix material" is meant a biomaterial found in natural mammalian tissues. Non-limiting examples of such physiologic matrix materials include: fibronectin, thrombospondin, glycosaminoglycans (GAGs) (e.g., hyaluronic acid, 6-sulfated chondroitin, dermatan sulfate, 4-sulfated chondroitin, or keratan sulfate), deoxyribonucleic acid (DNA), adhesive glycoproteins, and collagens (e.g., collagen I, collagen II, collagen III, collagen IV, collagen V, collagen VI, or collagen XVIII).

[0068] Collagen provides tensile strength to most tissues, and multiple collagen fibrils of about 100 nm in diameter combine to generate strong, coiled helical fibers of about 10 pm in diameter. The biomechanical function of certain tissue architectures is imparted via the directional arrangement of collagen fibers. In some embodiments, the input material includes collagen fibrils. Input materials including collagen fibrils can be used to create a fibrous structure that forms a tissue architecture. By adjusting the diameter of the fibrous structure, the directionality of the collagen fibrils can be controlled, thereby directing the aggregation of the collagen fibrils in a desired manner.

[0069] In embodiments, the plurality of cells is selected from, or consists of, the group including a single cell suspension, a cell aggregate, a cell spheroid, a cell organoid, or combinations thereof. The flow material according to embodiments of the present application can incorporate any mammalian cell type, including but not limited to stem cells (e.g., embryonic stem cells, adult stem cells, induced pluripotent stem cells), germ cells, endodermal cells (e.g., lung cells, liver cells, pancreatic cells, gastrointestinal cells, or genitourinary tract cells), mesodermal cells (e.g., kidney cells, bone cells, muscle cells, endothelial cells, or heart cells), and ectodermal cells (skin cells, nervous system cells, pituitary cells, or eye cells), stem cell-derived cells, or any combination thereof.

[0070] For example, the flowing material can include cells from endocrine and exocrine glands, including pancreas (alpha, beta, delta, epsilon, gamma), liver (hepatocytes, Kuppfer cells, stellate cells, sinusoidal endothelial cells, bile duct cells), thyroid (follicular cells), pineal gland (pinealocytes), pituitary gland (somatotrophs, lactotrophs, gonadotrophs, corticotrophs, and thyrotrophs), thymus (thymocytes, thymic epithelial cells, thymic stromal cells), adrenal gland (cortical cells, chromaffin cells), ovary (granulosa cells), testis (interstitial cells), gastrointestinal tract (enteroendocrine cells - intestine, stomach, pancreas), fibroblast cells, chondrocytes, meniscal fibrocartilage cells, bone marrow stromal (stem) cells, embryonic stem cells, mesenchymal stem cells, induced pluripotent stem cells, differentiated stem cells, tissue-derived cells, smooth muscle cells, skeletal muscle cells, cardiac muscle cells, epithelial cells, endothelial cells, myoblasts, chondroblasts, osteoblasts, osteoclasts, and any combination thereof.

[0071] The cells can be obtained from a donor of the same species as the recipient (allogeneic), from a different species than the recipient (xenogeneic), or from the recipient (autologous). Specifically, in embodiments, the cells can be obtained from a suitable donor, such as a human or an animal, or from the subject into which the cells are to be implanted. Mammalian species include, but are not limited to, humans, monkeys, dogs, cows, horses, pigs, sheep, goats, cats, mice, rabbits, and rats. In one embodiment, the cells are human cells. In other embodiments, the cells can be xenogeneic cells, for example, derived from an animal such as a dog, cat, horse, monkey, or any other mammal.

[0072] In some embodiments, the at least one biomaterial includes a plurality of cells that express / secret one or more endogenous bioactive agents (e.g., insulin, glucagon, ghrelin, pancreatic polypeptide, Factor VII, Factor VIII, Factor IX, alpha-1 antitrypsin, angiogenic factors, growth factors, hormones, antibodies, enzymes, proteins, exosomes, etc.). Endogenous bioactive agents discussed herein include those agents that are naturally produced by cells in a biological environment (e.g., insulin release in response to elevated glucose concentrations). Endogenous bioactive agents can constitute therapeutic agents in the context of the present disclosure.

[0073] In some embodiments, the flowing material can include genetically engineered cells that secrete a particular factor. As discussed above, multiple cells in embodiments can include engineered cells (e.g., genetically engineered cells) that secrete a particular factor, which is within the scope of the present disclosure. Cells can also be from established cell culture lines, or can be cells that have been genetically engineered and / or manipulated to achieve a desired genotype or phenotype. In some embodiments, tissue pieces can also be used, which can provide multiple different cell types within the same structure.

[0074] Genetic engineering techniques suitable for use in the present disclosure can include, but are not limited to: recombinant DNA (rDNA) techniques (Stryjewska et al., Pharmacologial Reports. 2013; 65: 1075); cell engineering based on the use of targeted nucleases (e.g., meganucleases, zinc-finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs), clustered regularly interspaced short palindromic repeat sequence-associated nuclease Cas9 (CRISPR-Cas9), etc.) (Lim et al., Nature Communications. 2020; 11: 4043; Stoddard BL, Structure. 2011; 19(1): 7-15; Gaj et al., Trends Biotechnol. 2013; 31(7): 397-405; Hsu et al., Cell. 2014; 157(6): 1262; Miller et al., Nat Biotechnol. 2010; 29(2): 143-148); cell engineering based on the use of site-specific recombination using recombinase systems (e.g., Cre-Lox) (Osborn et al., Mol Ther. 2013; 21(6): 1151-1159; Hockemeyer et al., Nat Biotechnol. 2009; 27(9): 851-857; Uhde-Stone et al., RNA. 2014; 20(6): 948-955; Ho et al., Nucleic Acids Res. 2015; 43(3): e17; Sengupta et al., Journal of Biological Engineering. 2017; 11(45): 1-9), etc. In some embodiments, some combination of the above techniques for cell engineering can be used.

[0075] The present disclosure encompasses engineered cells capable of producing one or more therapeutic agents including, but not limited to, proteins, peptides, nucleic acids (e.g., DNA, RNA, mRNA, siRNA, miRNA, nucleic acid analogs), peptide nucleic acids, aptamers, antibodies or fragments or portions thereof, antigens or epitopes, hormones, hormone antagonists, growth factors or recombinant growth factors and fragments and variants thereof, cytokines, enzymes, antibiotics or antimicrobial compounds, anti-inflammatory agents, antifungal agents, antiviral agents, toxins, prodrugs, small molecules, drugs (e.g., pharmaceuticals, dyes, amino acids, vitamins, antioxidants), or any combination thereof. EMBODIMENTS

[0076] A cumulative dataset of 660 images with cell, aggregate, and bead annotations was collected from pre-print, during-print, and post-print stages of the bioprinting pipeline. The model architecture chosen was an adapted Torchvision implementation of the Mask R-CNN instance segmentation model with a ResNet-50-FPN backbone and pre-trained on the Common Images in COntext (COCO) dataset. Images were fed into the model in a tiled fashion by sliding a moving window over each image. Various dynamic data augmentation methods were utilized to improve the model’s generalizability, including random rotations, scaling, translations, color jitter changes in brightness, contrast, saturation, and hue, addition of Gaussian noise, random perspective transformation, random CLAHE, and gamma transformation, and random image sharpening and blurring. For model optimization, random gradient descent with a step decay strategy was used. The method’s hyperparameters were adjusted and fine-tuned according to the downstream task. Figure 21A 、 Figure 21D 、 Figure 21G 、 Figure 21J and Figure 21C Sample segmentation and quantification results for each printing stage are shown. The following are descriptions of these results.

[0077] • Figure 21F (Pre-print cells): These figures show examples of print cell quantification with aggregates, cells, and debris outlined. Cell quality features are computed based on the annotations and stored. Figure 21I and Figure 21L are bar charts corresponding to Figures 6A to 6D and Figures 11A to 11C respectively, and show the value distribution of area, diameter, perimeter, roundness, smoothness, and compactness, which are the features discussed previously.

[0078] • Figures 22A to 22F Figures 23A to 23F Figures 24A to 24P Figures 25A to 25F Figure 26 Figures 22A to 22F Figure 22B Figure 22E Figure 22A Figure 22D Figures 23A to 23F (During-print cells): • These figures show cell quantification examples during printing, tracing out fibers, aggregates and cells using a camera trained on the print head nozzles. Cell mass features are computed based on the annotations and stored. FIG. 23B and 25E Bar plots corresponding to FIG. 23A and FIG. 23D are shown, and the value distributions for area, diameter, perimeter, roundness, smoothness and compactness, which are the features discussed earlier, are shown.

[0079] • FIGS. 24A-24P (Post-print cells): • For post-print measurements, the same aggregate and cell detection and morphology as for pre-print and during-print analysis is provided. In the post-print example, a measure of line density is also considered, defined as the expected number of aggregates per unit length of fiber. To avoid areas of complexity where fibers overlap or meet the image edges, “segment” and “midline” annotations are provided to allow for targeted measurement of aggregate density. FIG. 24B and FIG. 24C , FIG. 24F and FIG. 24G , FIG. 24J and FIG. 24K , and FIG. 24N and FIG. 24O Bar plots corresponding to FIG. 24A , FIG. 24D , FIG. 24I and FIG. 24M are shown, and the value distributions for area, diameter, perimeter, roundness, smoothness and compactness, which are the features discussed earlier, are shown, as well as core diameter.

[0080] • FIGS. 25A-25F (In-print beads): • These figures show in-print bead quantification examples, tracing out beads in the printed fibers. FIG. 25B and 25E Bar plots corresponding to FIG. 25A and FIG. 25D are shown, and the value distributions for area, diameter, perimeter, roundness, smoothness and compactness, which are the features discussed earlier, are shown.

[0081] • FIG. 26 (Software integration): • This figure shows the bead quantification model integrated into the print analysis software for Aspect. The software depicts bead quantification and relative distribution over the length of the print, as shown on the plots. Fiber diameter over the length of the print is also shown. Several video feeds of the print process, taken from the perspective of the nozzles and the print surface, as well as a representation of the tool path that defines the print pattern, can be seen.

[0082] All patents and non-patent literature cited in this specification are hereby incorporated by reference in their entirety and for all purposes.

[0083] Aspects of the application are set out in the following clauses: Clause 1. A bioprinting system comprising: a bioprinter, optionally comprising one or more reservoirs, a print head, and a print surface onto which a bioprint is dispensed; an optical system for collecting imaging data of a plurality of cells within a material stream comprising a cross-linkable material before, during, and / or after the plurality of cells pass through the bioprinting system; and a computing system for processing the imaging data to determine a number, distribution, concentration, and / or morphology of the plurality of cells; wherein the computing system comprises a machine learning system that: compares a number, distribution, concentration, and / or morphology of the plurality of cells before, during, and / or after passing through the bioprinting system to a user-established cell morphology corresponding to physical properties of cells and / or cell aggregates within a predetermined tolerance, and / or compares a number, distribution, concentration, and / or morphology of the plurality of cells at one stage before, during, and / or after the plurality of cells pass through the bioprinting system to a number, distribution, concentration, and / or morphology of the plurality of cells at another stage before, during, and / or after the plurality of cells pass through the bioprinting system, and records the number, distribution, concentration, and / or morphology of the plurality of cells before, during, and / or after passing through the bioprinting system, and a result of the comparison.

[0084] Clause 2. The system of clause 1, wherein the cross-linkable material is cross-linked to form a printed microcapsule or fiber; preferably wherein the cross-linkable material is cross-linked to form a printed fiber.

[0085] Clause 3. The system of clause 1, wherein the machine learning system performs image segmentation on the imaging data to determine the cell morphology.

[0086] Clause 4. The system of clause 3, wherein the machine learning system comprises a neural network selected from the group consisting of a convolutional neural network (CNN), a fully convolutional network (FCN), a region-based CNN (R-CNN), a Mask R-CNN, a model based on You Only Look Once (YOLO), a transformer-based instance segmentation model, a flow estimation network selected from the group consisting of the FlowNet series, a recurrent affine transformation (RAFT), and a self-learning multi-frame unsupervised RAFT with full-image warping (SMURF) to perform the image segmentation on the imaging data.

[0087] Clause 5. The system of clause 3 or clause 4, wherein the image segmentation comprises a panoptic segmentation and / or a semantic segmentation and / or an instance segmentation.

[0088] Clause 6. The system of any preceding clause, wherein, in response to the comparison, the computing system adjusts one or more parameters of the bioprinting system to change a flow of the plurality of cells through the print head.

[0089] Clause 7. The system of any preceding clause, wherein the image data comprises a plurality of image frames of a portion of a printed fiber, and the computing system assembles the plurality of image frames to provide an image of an entire length of the printed fiber.

[0090] Clause 8. The system of clause 6, wherein the computing system uses the image of the entire length of the printed fiber to obtain a measure of a number of cells inside the printed fiber.

[0091] Clause 9. The system of clause 7, wherein the computing system uses the measure of the number of cells inside the printed fiber, and / or a comparison of the number, distribution, concentration, and / or morphology of the cells before, during, and / or after the plurality of cells pass through the bioprinting system to calculate a therapeutic dose.

[0092] Clause 10. The system of any preceding clause, wherein the print head has one or more transparent channels, wherein the optical system collects imaging data of the plurality of cells through at least one of the one or more transparent channels, preferably wherein the print head comprises a transparent nozzle or dispensing channel.

[0093] Clause 11. A system according to any of the preceding clauses, wherein the optical system comprises one or more imaging devices selected from the group consisting of microscopes, still cameras and video cameras, the optical system being positioned at different locations in the bioprinting system to collect the imaging data before, during and / or after the plurality of cells pass through the bioprinting system.

[0094] Clause 12. The system according to Clause 10, wherein the optical system includes one or more imaging devices positioned at each of the input reservoir, the print head, and the print surface.

[0095] Clause 13. The system according to any of the preceding clauses, wherein the machine learning system distinguishes between focused imaging data and out-of-focus imaging data, and performs the comparison and recording on the focused imaging data.

[0096] Clause 14. The system according to any of the preceding clauses, wherein the machine learning system distinguishes between focused imaging data and out-of-focus imaging data, and improves focus on the out-of-focus imaging data to provide further focused imaging data, to perform the comparison and recording of both the focused imaging data and the further focused imaging data.

[0097] Clause 15. A system according to any of the preceding clauses, wherein the imaging data comprises a series of Z-stack images of a printed optical fiber, wherein the machine learning model detects focal cells and aggregates in different Z-stack layers of the printed optical fiber, such that focal changes of the optical system used to image the different Z-stack layers quantize the focal cells and aggregates in the different Z-stack layers.

[0098] Clause 16. The system according to Clause 1, wherein the printhead comprises a printhead selected from the group consisting of an extrusion printhead, a co-extrusion printhead, an inkjet printhead, an electro-writing printhead, and a microfluidic printhead.

[0099] Clause 17. The system according to any of the preceding clauses further includes determining the quality of the plurality of cells in response to determination of quantity, distribution, concentration and / or morphology.

[0100] Clause 18. The system according to any of the preceding clauses, wherein the morphology comprises at least one characteristic selected from the group consisting of cell count, area, perimeter, ellipsoidal volume, roundness, smoothness, compactness, circle diameter, average diameter, minimum diameter, maximum diameter, average strength, and cell density in the fiber.

[0101] Clause 19. A bioprinting method, comprising: imaging data of the plurality of cells within a stream of material comprising cross-linkable material before, during, and / or after the plurality of cells pass through a bioprinting system, the bioprinting system comprising a bioprinter, optionally comprising one or more input reservoirs, a print head, and a print surface onto which the bioprint is dispensed; processing the imaging data with a computer to determine a number, distribution, concentration, and / or morphology of the plurality of cells, the processing comprising: comparing, with a machine learning system, the number, distribution, concentration, and / or morphology of the plurality of cells at one of before, during, and / or after cross-linking in the bioprinting system to a user-established cell morphology corresponding to physical properties of cells and / or cell aggregates within a predetermined tolerance, and / or comparing, with the machine learning system, the number, distribution, concentration, and / or morphology of the plurality of cells at one of before, during, and / or after cross-linking in the bioprinting system to the number, distribution, concentration, and / or morphology of the plurality of cells at another of before, during, and / or after the plurality of cells pass through the bioprinting system, and recording, with the machine learning system, the number, distribution, concentration, and / or morphology of the plurality of cells before, during, and / or after the plurality of cells pass through the bioprinting system, and a result of the comparing.

[0102] Clause 20. The method of Clause 19, wherein the cross-linkable material is cross-linked to form printed microcapsules or fibers; preferably wherein the cross-linkable material is cross-linked to form printed fibers.

[0103] Clause 21. The method of Clause 19, further comprising performing image segmentation on the imaging data with the machine learning system to determine the cell morphology.

[0104] Clause 22. The method of Clause 19, wherein the machine learning system comprises a neural network selected from the group consisting of a convolutional neural network (CNN), a region-based CNN (R-CNN), a Mask R-CNN, a you only look once (YOLO)-based model, a transformer-based instance segmentation model, a flow-estimation network selected from the group consisting of the FlowNet series, a recurrent affine transformation (RAFT), and a self-learning multi-frame unsupervised RAFT (SMURF) with full-image warping, to perform the image segmentation on the imaging data.

[0105] Clause 23. The method of Clause 21 or Clause 22, wherein the image segmentation comprises panoramic segmentation and / or semantic segmentation and / or instance segmentation.

[0106] Clause 24. The method of any one of clauses 19-23, further comprising adjusting, with the computer, one or more parameters of the bioprinting system to change a flow of the cross-linkable material through the print head in response to one or both of the comparisons.

[0107] Clause 25. The method of any one of clauses 19-24, wherein the image data comprises a plurality of image frames of a portion of a printed fiber, the method further comprising assembling, with the computer, the plurality of image frames to provide an image of an entire length of the printed fiber.

[0108] Clause 26. The method of clause 25, further comprising obtaining, with the computer, a measure of a number of cells inside the printed fiber using the image of the entire length of the printed fiber.

[0109] Clause 27. The method of clause 26, further comprising calculating a therapeutic dose in response to the obtaining and / or comparing of the number, distribution, concentration, mass, and / or morphology of the cells before, during, and / or after cross-linking in the bioprinting system.

[0110] Clause 28. The method of clause 19, wherein the print head has one or more transparent channels, the method further comprising collecting, with the optical system, imaging data of the plurality of cells through at least one of the one or more transparent channels, preferably wherein the print head comprises a transparent nozzle or dispensing channel.

[0111] Clause 29. The method of clause 19, wherein collecting the imaging data comprises providing one or more imaging devices in the optical system selected from the group consisting of a microscope, a still camera, and a video camera, and positioning the optical system at different locations in the bioprinting system to collect the imaging data before, during, and / or after the plurality of cells pass through the bioprinting system.

[0112] Clause 30. The method of clause 29, wherein the positioning comprises one or more imaging devices positioned at each of the input reservoir, the print head, and the print surface.

[0113] Clause 31. The method of any one of clauses 19-30, wherein the machine learning system distinguishes in-focus imaging data and out-of-focus imaging data, and performs the comparing and recording on the in-focus imaging data.

[0114] Clause 32. The method of any of clauses 19-31, wherein the machine learning system distinguishes in-focus imaging data and out-of-focus imaging data and improves focus on the out-of-focus imaging data to provide further in-focus imaging data to perform the comparison and recording of both the in-focus imaging data and the further in-focus imaging data.

[0115] Clause 33. The method of any of clauses 19-32, wherein the imaging data comprises a series of Z-stack images of a printed fiber, wherein the machine learning model detects in-focus cells and aggregates in different Z-stack layers of the printed fiber such that a change in focus of the optical system used to image different Z-stack layers enables quantification of the in-focus cells and aggregates in the different Z-stack layers.

[0116] Clause 34. The method of any of clauses 19-33, wherein the print head comprises a print head selected from the group consisting of an extrusion print head, a co-extrusion print head, an inkjet print head, and a microfluidic print head.

[0117] Clause 35. The method of any of clauses 19-34, further comprising determining a quality of the plurality of cells in response to the determination of quantity, distribution, concentration, and / or morphology.

[0118] Clause 36. The method of any of clauses 19-35, wherein the morphology comprises at least one property selected from the group consisting of cell count, area, perimeter, ellipsoid volume, roundness, smoothness, compactness, circular diameter, average diameter, minimum diameter, maximum diameter, average intensity, and cell density in a fiber.

Claims

1. A bioprinting system, comprising: A bioprinter may optionally include one or more reservoirs, a printhead, and a printing surface on which the bioprinter is dispensed; An optical system for collecting imaging data of the plurality of cells within a material flow containing a crosslinkable material before, during, and / or after the plurality of cells pass through the bioprinting system; as well as A computing system for processing the imaging data to determine the number, distribution, concentration, and / or morphology of the plurality of cells; The computing system mentioned above includes a machine learning system, which: The number, distribution, concentration, and / or morphology of the plurality of cells before, during, and / or after passing through the bioprinting system are compared with a user-defined cell morphology corresponding to the physical properties of cells and / or cell aggregates within a predetermined tolerance. The quantity, distribution, concentration, and / or morphology of the plurality of cells in one phase before, during, and / or after the plurality of cells pass through the bioprinting system are compared with the quantity, distribution, concentration, and / or morphology of the plurality of cells in another phase before, during, and / or after the plurality of cells pass through the bioprinting system. The quantity, distribution, concentration, and / or morphology of the plurality of cells are recorded before, during, and / or after the plurality of cells pass through the bioprinting system, as well as the results of the comparison.

2. The system of claim 1, wherein the crosslinkable material is crosslinked to form printed microcapsules or fibers; preferably, wherein the crosslinkable material is crosslinked to form printed fibers.

3. The system of claim 1, wherein the machine learning system performs image segmentation on the imaging data to determine the cell morphology.

4. The system of claim 3, wherein the machine learning system comprises a neural network selected from the group consisting of a convolutional neural network (CNN), a fully convolutional network (FCN), a region-based CNN (R-CNN), a Mask R-CNN, a YOLO-based model, a transformer-based instance segmentation model, a flow estimation network selected from the group consisting of flow estimation networks from the FlowNet family, a recurrent full-pair field transform (RAFT), and a self-learning multi-frame unsupervised RAFT (SMURF) with full image distortion, to perform the image segmentation on the imaging data.

5. The system according to claim 3 or claim 4, wherein the image segmentation includes panoramic segmentation and / or semantic segmentation and / or instance segmentation.

6. The system according to any of the preceding claims, wherein, In response to the comparison, the computing system adjusts one or more parameters of the bioprinting system to alter the flow of the plurality of cells through the printhead.

7. The system according to any of the preceding claims, wherein the image data comprises a plurality of image frames of a portion of the printed optical fiber, and the computing system assembles the plurality of image frames to provide an image of the entire length of the printed optical fiber.

8. The system of claim 6, wherein the computing system uses the image of the entire length of the printed fiber to obtain a measurement of the number of cells inside the printed fiber.

9. The system of claim 7, wherein the computing system uses the measured number of cells inside the printed fiber, and / or a comparison of the number, distribution, concentration, and / or morphology of the cells before, during, and / or after the plurality of cells pass through the bioprinting system to calculate the therapeutic dose.

10. The system according to any of the preceding claims, wherein the printhead has one or more transparent channels, wherein the optical system collects imaging data of the plurality of cells through at least one of the one or more transparent channels, preferably wherein the printhead includes a transparent nozzle or a dispensing channel.

11. The system according to any of the preceding claims, wherein the optical system comprises one or more imaging devices selected from the group consisting of microscopes, still cameras and video cameras, the optical system being positioned at different locations in the bioprinting system to collect the imaging data before, during and / or after the plurality of cells pass through the bioprinting system.

12. The system of claim 10, wherein the optical system comprises one or more imaging devices positioned at each of the input reservoir, the print head, and the print surface.

13. The system according to any of the preceding claims, wherein the machine learning system distinguishes between focused imaging data and out-of-focus imaging data, and performs the comparison and recording on the focused imaging data.

14. The system according to any of the preceding claims, wherein the machine learning system distinguishes between focused imaging data and out-of-focus imaging data, and improves focus on the out-of-focus imaging data to provide further focused imaging data, to perform the comparison and recording of both the focused imaging data and the further focused imaging data.

15. The system according to any of the preceding claims, wherein the imaging data comprises a series of Z-stack images of a printed optical fiber, wherein the machine learning model detects focal cells and aggregates in different Z-stack layers of the printed optical fiber, such that focal changes of the optical system used to image the different Z-stack layers quantize the focal cells and aggregates in the different Z-stack layers.

16. The system of claim 1, wherein the printhead comprises a printhead selected from the group consisting of an extrusion printhead, a co-extrusion printhead, an inkjet printhead, an electro-writing printhead, and a microfluidic printhead.

17. The system according to any of the preceding claims further includes determining the mass of the plurality of cells in response to determination of quantity, distribution, concentration and / or morphology.

18. The system according to any of the preceding claims, wherein the morphology comprises at least one characteristic selected from the group consisting of cell count, area, perimeter, ellipsoidal volume, roundness, smoothness, compactness, circle diameter, average diameter, minimum diameter, maximum diameter, average strength, and cell density in the fiber.

19. A bioprinting method, comprising: Imaging data of the multiple cells, including a crosslinkable material, are collected before, during, and / or after they pass through a bioprinting system, which includes a bioprinter and optionally includes one or more input reservoirs, a printhead, and a printing surface on which the bioprinter is dispensed. Using a computer to process the imaging data to determine the number, distribution, concentration, and / or morphology of the plurality of cells, the processing includes: The machine learning system is used to compare the number, distribution, concentration, and / or morphology of the multiple cells before, during, and / or after passing through the bioprinting system with a user-defined cell morphology corresponding to the physical properties of cells and / or cell aggregates within a predetermined tolerance. The machine learning system is used to compare the quantity, distribution, concentration, and / or morphology of the plurality of cells in one stage before, during, and / or after cross-linking in the bioprinting system with the quantity, distribution, concentration, and / or morphology of the plurality of cells in another stage before, during, and / or after the plurality of cells pass through the bioprinting system. The machine learning system is used to record the number, distribution, concentration, and / or morphology of the plurality of cells before, during, and / or after the plurality of cells pass through the bioprinting system, as well as the results of the comparison.

20. The method of claim 19, wherein the crosslinkable material is crosslinked to form printed microcapsules or fibers; preferably, wherein the crosslinkable material is crosslinked to form printed fibers.

21. The method of claim 19, further comprising using the machine learning system to perform image segmentation on the imaging data to determine the cell morphology.

22. The method of claim 19, wherein the machine learning system comprises a neural network selected from the group consisting of a convolutional neural network (CNN), a region-based CNN (R-CNN), a Mask R-CNN, a YOLO-based model, a transformer-based instance segmentation model, a flow estimation network selected from the group consisting of flow estimation networks from the FlowNet family, a recurrent full-pair field transform (RAFT), and a self-learning multi-frame unsupervised RAFT (SMURF) with full image distortion, to perform the image segmentation on the imaging data.

23. The method of claim 21 or claim 22, wherein the image segmentation includes panorama segmentation and / or semantic segmentation and / or instance segmentation.

24. The method according to any one of claims 19 to 23, further comprising, in response to one or both of the comparisons, using the computer to adjust one or more parameters of the bioprinting system to alter the flow of the crosslinkable material through the printhead.

25. The method of any one of claims 19 to 24, wherein the image data comprises a plurality of image frames of a portion of the printed optical fiber, the method further comprising assembling the plurality of image frames using the computer to provide an image of the entire length of the printed optical fiber.

26. The method of claim 25, further comprising using the computer to obtain a measurement of the number of cells inside the printed fiber using the image of the entire length of the printed fiber.

27. The method of claim 26, further comprising calculating a therapeutic dose in response to the acquisition and / or comparison of the number, distribution, concentration, quality and / or morphology of the cells before, during and / or after crosslinking in the bioprinting system.

28. The method of claim 19, wherein the printhead has one or more transparent channels, the method further comprising using the optical system to collect imaging data of the plurality of cells through at least one of the one or more transparent channels, preferably wherein the printhead includes a transparent nozzle or a dispensing channel.

29. The method of claim 19, wherein collecting the imaging data comprises providing one or more imaging devices selected from the group consisting of a microscope, a still camera, and a video camera in the optical system, and positioning the optical system at different locations in the bioprinting system to collect the imaging data before, during, and / or after the plurality of cells pass through the bioprinting system.

30. The method of claim 29, wherein the positioning includes one or more imaging devices positioned at each of the input reservoir, the print head, and the print surface.

31. The method according to any one of claims 19 to 30, wherein the machine learning system distinguishes between focused imaging data and out-of-focus imaging data, and performs the comparison and recording on the focused imaging data.

32. The method according to any one of claims 19 to 31, wherein the machine learning system distinguishes between focused imaging data and out-of-focus imaging data, and improves the focus on the out-of-focus imaging data to provide further focused imaging data, to perform the comparison and recording of both the focused imaging data and the further focused imaging data.

33. The method according to any one of claims 19 to 32, wherein the imaging data comprises a series of Z-stack images of a printed optical fiber, wherein the machine learning model detects focal cells and aggregates in different Z-stack layers of the printed optical fiber, such that focal changes of the optical system used to image the different Z-stack layers quantize the focal cells and aggregates in the different Z-stack layers.

34. The method according to any one of claims 19 to 33, wherein the printhead comprises a printhead selected from the group consisting of an extrusion printhead, a co-extrusion printhead, an inkjet printhead, and a microfluidic printhead.

35. The method according to any one of claims 19 to 34, further comprising determining the quality of the plurality of cells in response to determination of quantity, distribution, concentration and / or morphology.

36. The method according to any one of claims 19 to 35, wherein the morphology comprises at least one characteristic selected from the group consisting of cell count, area, perimeter, ellipsoidal volume, roundness, smoothness, compactness, circle diameter, average diameter, minimum diameter, maximum diameter, average strength, and cell density in the fiber.

Images