Tissue sample quality evaluation method and device, analysis equipment and storage medium

Abstract

The invention relates to the field of biology, and discloses a tissue sample quality evaluation method and device, analysis equipment and a storage medium. The method comprises the following steps: acquiring basic information, a sample image and a sample video of a preset tissue sample; determining a first feature matrix representing the basic information; determining a second feature matrix representing the tissue sample by using a preset first extraction model according to the sample image; determining a third feature matrix representing the tissue sample by using a preset second extraction model according to the sample video; and determining a quality score of the tissue sample by using a preset scoring model according to the first feature matrix, the second feature matrix and the third feature matrix. Therefore, the quality score of the tissue sample can be automatically determined according to the basic information, the sample image and the sample video of the tissue sample through the trained scoring model, the quality score of the tissue sample does not need to be determined by human eyes, and the efficiency of determining the quality of the tissue sample can be improved.

Description

Technical Field

[0001] The present application relates to the biological field, and in particular to a method and device for evaluating the quality of tissue samples, an analysis device, and a storage medium. Background Art

[0002] Through a large number of experiments, it is found that the quality of tissue samples will affect the quality of organisms cultured from the tissue samples. Therefore, the quality of tissue samples is screened and only relatively high-quality tissue samples are used as raw materials for biological culture, which will greatly improve the quality of cultured organisms. In related technologies, the quality of tissue samples is usually observed and evaluated by researchers. However, the efficiency of determining the quality of tissue samples through manual observation is low.

[0003] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present application, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0004] In order to provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. The summary is not an extensive review, nor is it intended to identify key / critical components or delineate the scope of protection of these embodiments, but rather serves as a prelude to the detailed description that follows.

[0005] The embodiments of the present application provide a tissue sample quality evaluation method and apparatus, an analysis device, and a storage medium to improve the efficiency of determining the quality of a tissue sample.

[0006] An embodiment of the present application provides a method for evaluating the quality of a tissue sample, including: obtaining basic information, a sample image, and a sample video of a preset tissue sample; determining a first feature matrix characterizing the basic information; determining a second feature matrix characterizing the tissue sample based on the sample image using a preset first extraction model; determining a third feature matrix characterizing the tissue sample based on the sample video using a preset second extraction model; and determining a quality score of the tissue sample based on the first feature matrix, the second feature matrix, and the third feature matrix using a preset scoring model.

[0007] In the above-mentioned implementation, by obtaining characteristic information of tissue samples from sample images and sample videos of tissue samples, and combining with basic information of tissue samples, a more comprehensive situation of tissue samples can be obtained. At the same time, the model can learn to capture the potential laws, patterns and structures behind the data. Thus, the quality score of tissue samples can be automatically determined based on the basic information, sample images and sample videos of tissue samples through a trained scoring model, without the need for manual naked eye observation to determine the quality of tissue samples, which can improve the efficiency of determining the quality of tissue samples. At the same time, due to the subjective component of the user's judgment during manual observation, the quality of tissue samples determined by manual observation has a large gap between the quality of tissue samples with the same score. However, the quality gap between tissue samples with the same score can be reduced through a trained scoring model.

[0008] Optionally, determining a first characteristic matrix representing the basic information includes: searching for coding information corresponding to the basic information in a preset coding mapping table; and converting the coding information into the first characteristic matrix using a preset digital coding algorithm.

[0009] In the above implementation, the basic information can be easily converted into the coded information through the pre-set coding mapping table. The coded information is converted into the first feature matrix using a digital coding algorithm, which can greatly simplify the subsequent data processing and analysis steps, making the data easier to be processed by the scoring model.

[0010] Optionally, determining a second feature matrix characterizing the tissue sample according to the sample image using a preset first extraction model includes: inputting the sample image into the preset first extraction model to obtain a second feature matrix characterizing color features and freshness features of the tissue sample.

[0011] In the above embodiment, the first extraction model can pre-learn how to identify the color features and freshness features of tissue samples. By inputting the sample image into the preset first extraction model, the second feature matrix representing the color features and freshness features of the tissue sample can be automatically, simply and accurately obtained.

[0012] Optionally, the sample video includes multiple frames of video images; a third feature matrix characterizing the tissue sample is determined based on the sample video using a preset second extraction model, including: screening a preset number of video images from the video images as candidate images; superimposing each of the candidate images in a time sequence in a depth direction to form a sequence of candidate images; inputting the sequence of candidate images into the preset second extraction model to obtain a third feature matrix characterizing the cell activity and cell proportion characteristics of the tissue sample.

[0013] In the above embodiment, by extracting a preset number of video images to form a candidate image sequence, the amount of calculation of the subsequent second extraction model can be reduced while maintaining the video image features that can reflect the tissue sample. At the same time, the second extraction model can learn in advance how to identify the cell activity and cell percentage features of the tissue sample. By inputting the candidate image sequence into the preset second extraction model, the third feature matrix that characterizes the cell activity and cell percentage features of the tissue sample can be automatically, simply and accurately obtained.

[0014] Optionally, determining the quality score of the tissue sample using a preset scoring model based on the first feature matrix, the second feature matrix and the third feature matrix includes: superimposing the first feature matrix, the second feature matrix and the third feature matrix to obtain an alternative feature matrix; and inputting the alternative feature matrix into a preset scoring model to obtain the quality score of the tissue sample.

[0015] In the above embodiment, by superimposing the first characteristic matrix, the second characteristic matrix and the third characteristic matrix, various information that can characterize the tissue sample information can be integrated. The candidate characteristic matrix that integrates various information that can characterize the tissue sample information is input into a preset scoring model, and the quality score of the tissue sample can be automatically and comprehensively determined, without the need for manual scoring of the tissue sample through observation, which can improve the efficiency of determining the quality of the tissue sample.

[0016] Optionally, the tissue sample is a tissue sample used to culture organoids; the basic information includes: the sampling method, the time when the sample is placed in the preservation solution after leaving the body, the time when the sample arrives at the laboratory after leaving the body, the temperature of the sample when it arrives at the laboratory, the sample quality and whether the preservation solution in which the sample is located is expired.

[0017] In the above embodiment, basic information such as the sampling method, the time when the sample is placed in the preservation solution after being removed from the body, the time when the sample arrives at the laboratory after being removed from the body, the temperature of the sample when it arrives at the laboratory, the sample quality and whether the preservation solution in which the sample is stored is expired can affect whether the tissue sample can be successfully cultured into an organoid. Therefore, using the sampling method, the time when the sample is placed in the preservation solution after being removed from the body, the time when the sample arrives at the laboratory after being removed from the body, the temperature of the sample when it arrives at the laboratory, the sample quality and whether the preservation solution in which the sample is stored is expired as basic information in the quality assessment of tissue samples can enable a more comprehensive and accurate assessment of the quality of tissue samples.

[0018] Optionally, the scoring model is obtained by: acquiring sample data with score labels; the sample data includes: basic information of tissue samples, sample images and sample videos; inputting the sample data with score labels into a preset regression network for training to obtain the scoring model.

[0019] In the above embodiment, by inputting sample data with score labels into a preset regression network for training, the regression network can learn the rules between different sample data and scores, thereby obtaining a scoring model that gives corresponding scores for tissue samples.

[0020] An embodiment of the present application provides a tissue sample quality evaluation device, including: an acquisition module, used to acquire preset basic information, sample images and sample videos of tissue samples; a first vector determination module, used to determine a first feature matrix characterizing the basic information; a second vector determination module, used to determine a second feature matrix characterizing the tissue sample according to the sample image using a preset first extraction model; a third vector determination module, used to determine a third feature matrix characterizing the tissue sample according to the sample video using a preset second extraction model; and a scoring module, used to determine a quality score of the tissue sample according to the first feature matrix, the second feature matrix and the third feature matrix using a preset scoring model.

[0021] An embodiment of the present application provides an analysis device, including a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the above-mentioned tissue sample quality evaluation method.

[0022] In some embodiments, the storage medium stores computer executable instructions, and when the computer executable instructions are called and executed by the processor, the computer executable instructions enable the processor to implement the above-mentioned tissue sample quality assessment method.

[0023] The above general description and the following description are exemplary and explanatory only and are not intended to limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] One or more embodiments are exemplarily described by corresponding drawings, and these exemplary descriptions and drawings do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are considered similar elements, and the drawings do not constitute a scale limitation, and wherein:

[0025] Figure 1 is a schematic diagram of a tissue sample quality evaluation method provided in an embodiment of the present application;

[0026] Figure 2 is a schematic diagram of a sample image provided in an embodiment of the present application;

[0027] Figure 3 is a schematic diagram of a frame of an image in a sample video provided in an embodiment of the present application;

[0028] Figure 4is a schematic diagram of a tissue sample quality evaluation device provided in an embodiment of the present application;

[0029] Figure 5 It is a schematic diagram of an analysis device provided in an embodiment of the present application.

[0030] Reference numerals:

[0031] 1: acquisition module; 2: first vector determination module; 3: second vector determination module; 4: third vector determination module; 5: scoring module; 6: bus; 7: processor; 8: memory; 9: communication interface. DETAILED DESCRIPTION

[0032] In order to be able to understand the features and technical contents of the embodiments of the present application in more detail, the implementation of the embodiments of the present application is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present application. In the following technical description, for the convenience of explanation, a full understanding of the disclosed embodiments is provided through multiple details. However, one or more embodiments can still be implemented without these details. In other cases, to simplify the drawings, well-known structures and devices can be simplified for display.

[0033] The terms "first", "second", etc. in the specification and claims of the embodiments of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the embodiments of the present application described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions.

[0034] Unless otherwise stated, the term "plurality" means two or more.

[0035] The term "correspondence" may refer to an association relationship or a binding relationship. The correspondence between A and B means that there is an association relationship or a binding relationship between A and B.

[0036] Embodiment 1

[0037] The present application provides a method for evaluating the quality of tissue samples. Figure 1 As shown, Figure 1 This is a schematic diagram of a method for evaluating the quality of tissue samples provided in an embodiment of the present application, comprising:

[0038] Step S101, obtaining basic information, sample images and sample videos of preset tissue samples.

[0039] It can be understood that the sample video is a sample microscopic video, that is, a video shot of a tissue sample under a microscope.

[0040] In some embodiments, the tissue sample may be a tissue sample for culturing an organoid, or a tissue sample for culturing a plant. It is understood that the tissue sample may also be other tissue samples that are applicable to the solutions of the present application.

[0041] In some embodiments, the basic information of the tissue sample may be the storage information corresponding to the tissue sample. For example: in the case where the tissue sample is a tissue sample used to culture organoids, the storage information may be the sampling method, the time when the sample is placed in the preservation solution after being removed from the body, the time when the sample is removed from the body and arrives at the laboratory, the temperature at which the sample arrives at the laboratory, the sample quality, and whether the preservation solution in which the sample is located is expired, etc. Among them, the sampling method may be solid sample surgical sampling, solid sample endoscopic / puncture sampling, liquid sample drainage sampling, liquid sample puncture sampling, etc. For another example: in the case where the tissue sample is a tissue sample used to culture plants, the storage information may be the sampling site, the time when the sample is placed in the preservation solution after being removed from the body, the time when the sample is removed from the body and arrives at the laboratory, the temperature at which the sample arrives at the laboratory, the sample quality, and whether the preservation solution in which the sample is located is expired, etc. Among them, the sampling site may be the main stem, leaves, etc.

[0042] In some embodiments, the items of basic information of preset tissue samples can also be set by scientific researchers based on experience. For example: scientific researchers can set the items of basic information including the reaction characteristics of the cell tissue suspension of the sample tissue and some reagents. Scientific researchers can set the items of basic information including storage information, transportation conditions, etc. For another example: scientific researchers can set the items of basic information including storage information, sampling body status, etc. Among them, the sampling body status is the status of the sampled body, and the sampling body status can be described by good status, poor status, etc.

[0043] It can be understood that any information that can reflect or describe relevant information of a tissue sample can be used as basic information.

[0044] Exemplarily, the tissue sample is a tissue sample for culturing organoids. The basic information of the tissue sample for culturing organoids is shown in the following table. The sampling method of tissue sample 1 is liquid sample drainage sampling. The time from the sample being removed from the body to being placed in the preservation solution is 25 minutes. The time from the sample being removed from the body to arriving at the laboratory is 6 hours. The temperature of the sample when it arrives at the laboratory is 7.8 degrees Celsius. The mass of the sample is 0.253 grams, and the preservation solution is not expired.

[0045]

[0046] Table 1

[0047] In some embodiments, the tissue sample may be photographed by a first imaging device or a first imaging apparatus to obtain a sample image.

[0048] Example 1: A first imaging device photographs a tissue sample to obtain a sample image, and then the first imaging device sends the sample image to an analysis device, which receives the sample image.

[0049] Example 2: The analysis device is integrated with a first imaging device, and the imaging device of the analysis device photographs the tissue sample to obtain a sample image.

[0050] In some embodiments, the tissue sample may be photographed by a second imaging device or a second imaging apparatus to obtain a sample video.

[0051] Example 1: The second imaging device photographs the tissue sample to obtain a sample video, and then the second imaging device sends the sample video to the analysis device, which receives the sample video.

[0052] Example 2: The analysis device is integrated with a second imaging device, and the imaging device of the analysis device photographs the tissue sample to obtain a sample video.

[0053] In the above embodiment, the first imaging device and the second imaging device may be integrated into the analysis device at the same time, the first imaging device may take a sample image, and the second imaging device may take a sample video.

[0054] In the above embodiment, the first imaging device or the first imaging apparatus may be a camera, and the second shooting device or the second imaging apparatus may be a microscope shooting device. Figure 2 and Figure 3 shown. Figure 2 A schematic diagram of a sample image taken by a first imaging device or a first imaging apparatus, such as Figure 2 It can be seen that the sample image of the tissue sample photographed by the camera can reflect the overall appearance of the tissue sample. Figure 3 As shown, the sample video shot by the microscope can show the cells of the tissue sample, thereby reflecting the cell activity of the tissue sample. In this way, shooting tissue samples by different types of shooting equipment can reflect the characteristics of different aspects of the tissue sample, thereby more comprehensively reflecting the characteristics of the tissue sample, and thus making the quality score of the tissue sample more objective and accurate.

[0055] Step S102: determine a first characteristic matrix representing basic information.

[0056] In some embodiments, step S102 may include: searching for coding information corresponding to the basic information in a preset coding mapping table; and converting the coding information into a first feature matrix using a preset digital coding algorithm.

[0057] Among them, the coding mapping table can be written by scientific researchers according to actual usage requirements.

[0058] For example, in combination with Table 2, Table 2 is an example table in which the information of the sampling method and the information of whether the preservation solution is expired in Table 1 are replaced with coding information. As shown in Table 2, the coding information corresponding to the sampling method of liquid sample drainage sampling is 1 in the preset coding mapping table, and the coding information corresponding to the sampling method of solid sample surgical operation sampling is 2 in the preset coding mapping table.

[0059] In the preset coding mapping table, find out whether the preservation solution is expired. If it is not expired, the corresponding coding information is 1.

[0060]

[0061] Table 2

[0062] The digital coding algorithm may be binary coding.

[0063] It is understandable that engineers may also choose other digital encoding algorithms, which are not limited here.

[0064] Step S103: Determine a second feature matrix representing the tissue sample according to the sample image using a preset first extraction model.

[0065] In some embodiments, step S103 includes: inputting the sample image into a preset first extraction model to obtain a second feature matrix characterizing the color features and freshness features of the tissue sample.

[0066] In the above embodiment, the preset first extraction model can be obtained in the following manner: the candidate image with color feature and freshness feature labels is input into a preset 2D convolution model for training to obtain the first extraction model.

[0067] Among them, the 2D convolutional model can use the first 15 layers of resnet_34 to extract features of the sample image.

[0068] Step S104: Determine a third feature matrix representing the tissue sample according to the sample video using a preset second extraction model.

[0069] In some embodiments, the sample video includes multiple frames of video images; step S104 may include: screening a preset number of video images from the multiple frames of video images as candidate images; superimposing the candidate images in the depth direction according to time sequence to form a sequence of candidate images; inputting the sequence of candidate images into a preset second extraction model to obtain a third feature matrix characterizing the cell activity and cell proportion characteristics of the tissue sample.

[0070] In the above embodiment, screening a preset number of video images from the video images as candidate images may be: extracting a preset number of video images as candidate images at preset time intervals.

[0071] Among them, the preset time interval can be set according to the experience of scientific researchers. It can be understood that the preset time is the time between two frames of video images. Exemplarily, it is assumed that the preset number is 2, the preset time is a, and a is equal to 2b. Multiple frames of video images, for example: video image A, video image B, video image C and video image D. Among them, it is assumed that the time interval between every two frames of video images is b. Then, video image A and video image C are selected from the video images as alternative images.

[0072] Exemplarily, stacking the candidate images in the depth direction in chronological order to form a candidate image sequence can be: using the function torch.stack. images_tensor = torch.stack (put the first candidate image in the stack, dim = 1); images_tensor = torch.stack (put the second candidate image in the stack, dim = 1), and so on, until the last candidate image to be stacked is put in, and the candidate image sequence is obtained. Among them, the images_tensor after the last candidate image to be stacked is put in is the candidate image sequence. dim is the preset dimension.

[0073] In the above embodiment, the preset second extraction model can be obtained in the following manner: the candidate image with cell activity and cell ratio labels is input into a preset 3D convolution model for training to obtain the second extraction model.

[0074] Among them, the 3D convolutional model can use the first 15 layers of resnet_34_3D to extract features of the candidate image sequence.

[0075] Step S105 : determining the quality score of the tissue sample using a preset scoring model according to the first feature matrix, the second feature matrix, and the third feature matrix.

[0076] In some embodiments, step S105 may include: superimposing the first feature matrix, the second feature matrix and the third feature matrix to obtain an alternative feature matrix; inputting the alternative feature matrix into a preset scoring model to obtain a quality score of the tissue sample.

[0077] In an optional manner in the above embodiment, the first characteristic matrix, the second characteristic matrix and the third characteristic matrix are superimposed to obtain an alternative characteristic matrix by: adjusting the shapes of the first characteristic matrix, the second characteristic matrix and the third characteristic matrix to be consistent; and superimposing the adjusted first characteristic matrix, the second characteristic matrix and the third characteristic matrix.

[0078] In the above optional method, the shapes of the first characteristic matrix, the second characteristic matrix and the third characteristic matrix are adjusted to be consistent, and the PCA (Principal Component Analysis) algorithm can be used to process the first characteristic matrix, the second characteristic matrix and the third characteristic matrix respectively, and reduce the dimensions of the first characteristic matrix, the second characteristic matrix and the third characteristic matrix to a preset dimension.

[0079] In the above optional manner, the shapes of the first characteristic matrix, the second characteristic matrix and the third characteristic matrix are adjusted to be consistent, and the first characteristic matrix, the second characteristic matrix and the third characteristic matrix can be adjusted to specific dimensions using a reshape function. It can be understood that the reshape function is a function in MATLAB (a software) that transforms a specified matrix into a matrix of a specific dimension.

[0080] In the above optional manner, the shapes of the first feature matrix, the second feature matrix and the third feature matrix are adjusted to be consistent by using a convolution method, that is, the first feature matrix, the second feature matrix and the third feature matrix are respectively input into a preset convolution layer to obtain feature matrices with consistent shapes. The height of the feature matrix after the convolution layer can be calculated. The width of the feature matrix after the convolution layer can be calculated. Among them, Hout is the output height, Hin is the input height, P is the padding, K is the height of the convolution kernel, and S is the stride. Wout is the output width, and Win is the input width. Therefore, engineers can set S, P, and K according to their own usage needs, that is, they can adjust the width and height of the feature matrix to the desired width and height.

[0081] For example: Assume that the attribute of the feature matrix is ​​(3, 224, 224), where the first column is the number of channels of the feature matrix, the second column is the height of the feature matrix, and the third column is the width of the feature matrix. (3, 224, 224) needs to be adjusted to (1, 56, 56). By setting the convolution layer to in_channels 3, out_channels 1, kernel 4, stride 4, and P 0, the feature matrix with the attribute (3, 224, 224) can be adjusted to the feature matrix with the attribute (1, 56, 56). It is worth noting that each feature matrix has three attributes: the number of channels, height, and width. By setting the parameters in_channels and out_channels of the convolution layer, the number of channels of the feature matrix can be directly changed.

[0082] In the above optional manner, superimposing the adjusted first characteristic matrix, the second characteristic matrix and the third characteristic matrix may be directly adding the adjusted first characteristic matrix, the second characteristic matrix and the third characteristic matrix to obtain an alternative characteristic matrix.

[0083] In the above optional manner, the adjusted first characteristic matrix, the second characteristic matrix and the third characteristic matrix are superimposed by respectively assigning weights to the adjusted first characteristic matrix, the second characteristic matrix and the third characteristic matrix, and adding the first characteristic matrix, the second characteristic matrix and the third characteristic matrix according to the weights to obtain an alternative characteristic matrix.

[0084] In the above embodiment, the scoring model is obtained by: obtaining sample data with score labels; the sample data includes: basic information of the tissue sample, sample images and sample videos; inputting the sample data with score labels into a preset regression network for training to obtain the scoring model.

[0085] It is understandable that the regression network can use two fully connected layers. The preset regression network can also be linear regression, ridge regression, polynomial regression, etc., which are not limited here.

[0086] The sample data with score labels are input into a preset regression network for training, which can be: for each group of sample data with score labels: determine the fourth feature matrix representing the basic information of the tissue sample in the sample data. Determine the fifth feature matrix representing the color characteristics and freshness characteristics of the tissue sample based on the sample graphics in the sample data. Determine the sixth feature matrix representing the cell activity and cell ratio characteristics of the tissue sample based on the sample video in the sample data. Superimpose the fourth feature matrix, the fifth feature matrix and the sixth feature matrix to obtain a sample feature matrix; input the sample feature matrix with score labels into the preset regression network for training to obtain a scoring model.

[0087] Determining the fourth characteristic matrix of the basic information representing the tissue sample in the sample data may include searching for the coding information corresponding to the basic information in a preset coding mapping table; and converting the coding information into the fourth characteristic matrix using a preset digital coding algorithm.

[0088] Among them, determining the fifth feature matrix characterizing the color characteristics and freshness characteristics of the tissue sample based on the sample graphics in the sample data can be done by inputting the sample image into a preset first extraction model to obtain the fifth feature matrix characterizing the color characteristics and freshness characteristics of the tissue sample.

[0089] Among them, determining the sixth characteristic matrix characterizing the cell activity and cell proportion characteristics of the tissue sample based on the sample video in the sample data can be performed by screening a preset number of video images from the video images as images to be superimposed, superimposing the images to be superimposed in the depth direction according to the time sequence to form a superimposed image sequence, and inputting the superimposed image sequence into a preset second extraction model to obtain the sixth characteristic matrix characterizing the cell activity and cell proportion characteristics of the tissue sample.

[0090] Among them, when using, how to superimpose the first characteristic matrix, the second characteristic matrix and the third characteristic matrix, the fourth characteristic matrix, the fifth characteristic matrix and the sixth characteristic matrix are superimposed by the same sampling method to obtain the sample characteristic matrix.

[0091] For example, assume that 300 sets of sample data are obtained, and the 300 sets of sample data are sent to a preset annotation platform, and the annotation platform feeds back sample data with score labels. Among them, the annotation platform can manually mark the quality scores for the sample data. Set the loss function of the training scoring model to the mean square error loss function, set the training mode of the training scoring model to use the optimization algorithm RMSprop (Root Mean Square Propagation, adaptive learning rate) training, and the epoch is 100.

[0092] Exemplarily, researchers input the basic information of the tissue sample into the analysis device, and place the tissue sample into a preset imaging area in the analysis device. The first imaging device and the second imaging device are both capable of imaging items placed in the imaging area. The first imaging device of the analysis device images the placed tissue sample to obtain a sample image, and the second imaging device of the analysis device images the placed tissue sample to obtain a sample video. Then, the analysis device determines a first feature matrix representing the basic information; determines a second feature matrix representing the tissue sample based on the sample image using a preset first extraction model; determines a third feature matrix representing the tissue sample based on the sample video using a preset second extraction model; and determines the quality score of the tissue sample using a preset scoring model based on the first feature matrix, the second feature matrix, and the third feature matrix.

[0093] Embodiment 2

[0094] Combination Figure 4 As shown, an embodiment of the present application provides a quality evaluation device for a tissue sample, comprising: an acquisition module 1, a first vector determination module 2, a second vector determination module 3, a third vector determination module 4 and a scoring module 5. The acquisition module 1 is used to acquire the preset basic information, sample image and sample video of the tissue sample; the first vector determination module 2 is used to determine the first feature matrix representing the basic information; the second vector determination module 3 is used to determine the second feature matrix representing the tissue sample according to the sample image using a preset first extraction model; the third vector determination module 4 is used to determine the third feature matrix representing the tissue sample according to the sample video using a preset second extraction model; the scoring module 5 is used to determine the quality score of the tissue sample according to the first feature matrix, the second feature matrix and the third feature matrix using a preset scoring model.

[0095] Optionally, the first vector determination module 2 is used to determine the first characteristic matrix representing the basic information in the following manner: searching for the coding information corresponding to the basic information in a preset coding mapping table; and converting the coding information into the first characteristic matrix using a preset digital coding algorithm.

[0096] Optionally, the second vector determination module 3 is used to determine a second feature matrix characterizing the tissue sample according to the sample image using a preset first extraction model in the following manner: input the sample image into the preset first extraction model to obtain a second feature matrix characterizing the color characteristics and freshness characteristics of the tissue sample.

[0097] Optionally, the sample video includes multiple frames of video images; the third vector determination module 4 is used to determine a third feature matrix characterizing the tissue sample based on the sample video using a preset second extraction model in the following manner: screening a preset number of video images from the video images as candidate images; superimposing each candidate image in a time sequence in a depth direction to form a candidate image sequence; inputting the candidate image sequence into the preset second extraction model to obtain a third feature matrix characterizing the cell activity and cell proportion characteristics of the tissue sample.

[0098] Optionally, the scoring module 5 is used to determine the quality score of the tissue sample using a preset scoring model based on the first feature matrix, the second feature matrix and the third feature matrix in the following manner: superimposing the first feature matrix, the second feature matrix and the third feature matrix to obtain an alternative feature matrix; inputting the alternative feature matrix into a preset scoring model to obtain a quality score of the tissue sample.

[0099] Optionally, the tissue sample quality evaluation device also includes: a model training module, which is used to obtain a scoring model in the following manner: obtaining sample data with score labels; the sample data includes: basic information of the tissue sample, sample images and sample videos; inputting the sample data with score labels into a preset regression network for training to obtain a scoring model.

[0100] Embodiment 3

[0101] Combination Figure 5 As shown, an embodiment of the present application provides an analysis device, including a processor 7 and a memory 8. Optionally, the device may also include a communication interface 9 and a bus 6. Among them, the processor 7, the communication interface 9, and the memory 8 can communicate with each other through the bus 6. The communication interface 9 can be used for information transmission. The processor 7 can call the logic instructions in the memory 8 to execute the method for determining the characteristics of the organoid in the above embodiment.

[0102] In addition, the logic instructions in the memory 8 mentioned above can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.

[0103] The memory 8 is a computer-readable storage medium that can be used to store software programs and computer executable programs, such as program instructions / modules corresponding to the method in the embodiment of the present application. The processor 7 executes the program instructions / modules stored in the memory 8 to perform functional applications and data processing, that is, to implement the method for determining the characteristics of organoids in the above embodiment.

[0104] The memory 8 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 8 may include a high-speed random access memory and may also include a non-volatile memory.

[0105] By using the analysis device provided in the embodiment of the present application, by obtaining the characteristic information of the tissue sample from the sample image and sample video of the tissue sample, and combining it with the basic information of the tissue sample, a more comprehensive situation of the tissue sample can be obtained. At the same time, the model can learn to capture the potential laws, patterns and structures behind the data. Thus, the quality score of the tissue sample can be automatically determined based on the basic information, sample image and sample video of the tissue sample through the trained scoring model, thereby improving the efficiency of determining the quality of the tissue sample.

[0106] An embodiment of the present application provides a computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to execute the above-mentioned method for determining organoid characteristics.

[0107] An embodiment of the present application provides a computer program product, which includes a computer program stored on a storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the above-mentioned method for determining organoid characteristics.

[0108] The computer-readable storage medium mentioned above may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.

[0109] The technical solution of the embodiment of the present application can be embodied in the form of a software product, which is stored in a storage medium and includes one or more instructions for a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method of the embodiment of the present application. The aforementioned storage medium may be a non-transient storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, and other media that can store program codes, or a transient storage medium.

[0110] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0111] The above are only embodiments of the present application and are not intended to limit the scope of protection of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application. At the same time, the above embodiments can be combined with each other to form new embodiments without conflict.

Claims

1. A method for evaluating the quality of a tissue sample, characterized in that: include: Obtaining basic information, sample images and sample videos of preset tissue samples; Determining a first characteristic matrix representing the basic information; Determine a second feature matrix representing the tissue sample using a preset first extraction model according to the sample image; Determine a third feature matrix characterizing the tissue sample using a preset second extraction model according to the sample video; A quality score of the tissue sample is determined using a preset scoring model according to the first feature matrix, the second feature matrix, and the third feature matrix.

2. The method according to claim 1, characterized in that Determining a first characteristic matrix characterizing the basic information includes: Searching for the encoding information corresponding to the basic information in a preset encoding mapping table; The coding information is converted into a first characteristic matrix using a preset digital coding algorithm.

3. The method according to claim 1, characterized in that Determining a second feature matrix characterizing the tissue sample according to the sample image using a preset first extraction model includes: The sample image is input into a preset first extraction model to obtain a second feature matrix representing the color features and freshness features of the tissue sample.

4. The method according to claim 1, characterized in that: The sample video includes multiple frames of video images; Determining a third feature matrix characterizing the tissue sample according to the sample video using a preset second extraction model includes: Filtering a preset number of video images from the video images as candidate images; Superimposing the candidate images in a depth direction according to the time sequence to form a candidate image sequence; The candidate image sequence is input into a preset second extraction model to obtain a third feature matrix characterizing the cell activity and cell ratio characteristics of the tissue sample.

5. The method according to any one of claims 1 to 4, characterized in that: Determining the quality score of the tissue sample using a preset scoring model according to the first feature matrix, the second feature matrix, and the third feature matrix includes: Superimposing the first characteristic matrix, the second characteristic matrix and the third characteristic matrix to obtain an alternative characteristic matrix; The candidate feature matrix is ​​input into a preset scoring model to obtain a quality score of the tissue sample.

6. The method according to any one of claims 1 to 4, characterized in that: The tissue sample is a tissue sample used to culture organoids; the basic information includes: the sampling method, the time when the sample is placed in the preservation solution after leaving the body, the time when the sample arrives at the laboratory after leaving the body, the temperature of the sample when it arrives at the laboratory, the sample quality and whether the preservation solution in which the sample is placed is expired.

7. The method according to any one of claims 1 to 4, characterized in that: The scoring model is obtained in the following way: Obtain sample data with score labels; the sample data includes: basic information of tissue samples, sample images and sample videos; The sample data with score labels are input into a preset regression network for training to obtain the scoring model.

8. A tissue sample quality evaluation device, characterized in that: include: An acquisition module, used to acquire basic information, sample images and sample videos of preset tissue samples; A first vector determination module, used to determine a first characteristic matrix representing the basic information; A second vector determination module, configured to determine a second feature matrix representing the tissue sample according to the sample image using a preset first extraction model; A third vector determination module, configured to determine a third feature matrix representing the tissue sample according to the sample video using a preset second extraction model; A scoring module is used to determine the quality score of the tissue sample using a preset scoring model according to the first feature matrix, the second feature matrix and the third feature matrix.

9. An analysis device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the tissue sample quality assessment method according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores computer executable instructions. When the computer executable instructions are called and executed by a processor, the computer executable instructions enable the processor to implement the tissue sample quality assessment method according to any one of claims 1 to 7.

Images