Method, device and system for analyzing characteristics of blastocyst stage embryos

Abstract

The present application relates to the field of assisted reproductive technology, and provides a method, device and embryo analysis system for analyzing the characteristics of blastocyst embryos. Compared with the end-to-end AI model, the analysis process and characteristics are more visible. In the present application, the contour of the trophoblast mask obtained by the segmentation of the blastocyst embryo image is extracted to obtain the outer contour and the inner contour of the trophoblast; the center point of the outer contour of the trophoblast is obtained, and the outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set; the lines between each outer contour sampling point and the center point in the outer contour sampling point set are obtained, and the inner contour sampling point set is obtained based on the intersection between each line and the inner contour of the trophoblast; the distance between the outer contour sampling point and the inner contour sampling point of the same sampling order is used to obtain the outer and inner point pair distances corresponding to each sampling order; the outer and inner point pair distances corresponding to each sampling order are mapped to the coordinate system to obtain a waveform, and the estimated number of cells on the trophoblast is obtained according to the number of peaks.

Description

Technical Field

[0001] The present application relates to the field of assisted reproductive technology, and in particular to a method, device, embryo analysis system, storage medium and computer program product for analyzing characteristics of blastocyst stage embryos. Background Art

[0002] Assisted reproductive technology is the use of artificial methods to fertilize eggs and sperm in vitro and culture them in vitro through an embryo incubator, and finally transplant them into the mother's uterus for pregnancy and development. Before embryo transplantation, it is necessary to evaluate the developmental potential of the blastocyst that has developed to the fifth day before transplantation.

[0003] Currently, AI (Artificial Intelligence) models can be used to evaluate blastocysts. However, most AI models use an end-to-end prediction model, that is, input image-output prediction result, and the intermediate analysis process and features are invisible. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device, embryo analysis system, storage medium and computer program product for analyzing the characteristics of blastocyst stage embryos in order to address the above technical problems.

[0005] The present application provides a method for analyzing characteristics of blastocyst-stage embryos, the method comprising:

[0006] The contour of the trophoblast mask obtained by segmenting the blastocyst stage embryo image is extracted to obtain the trophoblast outer contour and the trophoblast inner contour;

[0007] Obtaining the center point of the outer contour of the trophoblast, and sampling the outer contour of the trophoblast multiple times to obtain an outer contour sampling point set;

[0008] Obtaining connecting lines between each outer contour sampling point in the outer contour sampling point set and the center point, and obtaining an inner contour sampling point set based on intersection points between each connecting line and the inner contour of the trophoblast;

[0009] Based on the distance between the outer contour sampling points and the inner contour sampling points in the same sampling order, the outer and inner point pair distances corresponding to each sampling order are obtained;

[0010] The distances between the outer and inner point pairs corresponding to each sampling order are mapped to a coordinate system to obtain a waveform, and the estimated number of cells on the trophoblast is obtained based on the number of peaks in the waveform.

[0011] In one embodiment, obtaining the center point of the outer contour of the trophoblast layer includes:

[0012] Acquire a zero-order spatial moment, a first-order spatial moment about the horizontal axis, and a first-order spatial moment about the vertical axis of an image area enclosed by the outer contour of the trophoblast;

[0013] The center point of the outer contour of the trophoblast is determined in the horizontal and vertical directions according to the ratio of the first-order spatial moment about the horizontal axis to the zero-order spatial moment and the ratio of the first-order spatial moment about the vertical axis to the zero-order spatial moment.

[0014] In one embodiment, the outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set, including:

[0015] The outer contour of the trophoblast is sampled multiple times in a manner of uniformly sampling in 360° azimuth angle to obtain an outer contour sampling point set;

[0016] or,

[0017] According to the sampling method of the Douglas-Peucker algorithm, the outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set.

[0018] In one embodiment, obtaining an estimated number of cells on the trophoblast according to the number of peaks in the waveform includes:

[0019] Obtaining the total number of peaks in the waveform and the width of each peak;

[0020] From the total number of peaks, the number of peaks whose widths do not fall within a preset interval is subtracted; the preset interval is determined according to the number of pixel units occupied by the cells;

[0021] The subtraction result was used as an estimate of the number of cells on the trophoblast.

[0022] In one embodiment, after obtaining the estimated number of cells on the trophoblast, the method further comprises:

[0023] The distribution of the cells on the trophoblast is quantified according to the perimeter of the outer contour of the trophoblast and the estimated number of the cells to obtain a quantitative distribution result.

[0024] In one embodiment, the method further comprises:

[0025] Extracting the contour of the inner cell mass mask in the blastocyst stage embryo image to obtain the inner cell mass contour;

[0026] Based on the inner cell mass outline, the perimeter and area of ​​the inner cell mass are obtained;

[0027] The circularity of the inner cell mass was obtained based on the perimeter and area of ​​the inner cell mass.

[0028] In one embodiment, the method further comprises:

[0029] Collect images of the embryo during the blastocyst stage development process, obtain images of the blastocyst stage embryo at multiple imaging moments, and obtain the inner cell mass contour corresponding to each imaging moment;

[0030] According to the inner cell mass contour corresponding to each imaging moment, the center point position of the inner cell mass at each imaging moment is obtained;

[0031] The center point position of the inner cell mass at each imaging moment is mapped to the coordinate system to obtain the movement trajectory of the inner cell mass during the blastocyst stage development.

[0032] In one embodiment, the method further comprises:

[0033] Collect images of embryos during the development of the blastocyst stage, and obtain images of blastocyst stage embryos at multiple imaging moments;

[0034] Extract the corresponding blastocyst cavity mask and trophoblast outer contour from the blastocyst stage embryo image corresponding to the same imaging moment;

[0035] According to the image area enclosed by the blastocyst cavity mask and the outer contour of the trophoblast layer corresponding to the same imaging moment, the blastocyst cavity proportion corresponding to each of the multiple imaging moments is obtained;

[0036] According to the proportion of the blastocyst cavity corresponding to each of the multiple imaging moments, the change trend of the blastocyst cavity is obtained.

[0037] In one embodiment, the method further comprises:

[0038] Acquiring a blastocyst cavity mask extracted from the blastocyst stage embryo image;

[0039] Obtaining the blastocyst cavity ratio according to the area of ​​the blastocyst cavity mask and the area of ​​the image region enclosed by the outer contour of the trophoblast;

[0040] Based on the blastocyst cavity ratio, determining whether the embryo is in blastocyst stage III or blastocyst stage IV;

[0041] Obtaining the center point of the outer contour of the trophoblast layer, and sampling the outer contour of the trophoblast layer multiple times to obtain an outer contour sampling point set, including: if the embryo is in the blastocyst stage III or blastocyst stage IV, obtaining the center point of the outer contour of the trophoblast layer, and sampling the outer contour of the trophoblast layer multiple times to obtain an outer contour sampling point set;

[0042] or,

[0043] The contour of the inner cell mass mask in the blastocyst stage embryo image is extracted, including: if the embryo is in the blastocyst stage III or blastocyst stage IV, the contour of the inner cell mass mask in the blastocyst stage embryo image is extracted.

[0044] In one embodiment, before obtaining the blastocyst cavity ratio according to the area of ​​the blastocyst cavity mask and the area of ​​the image region enclosed by the outer contour of the trophoblast, the method further includes:

[0045] Determining the area of ​​the image region enclosed by the outer contour of the trophoblast according to the number of pixels within the outer contour of the trophoblast;

[0046] or,

[0047] The area of ​​the image region enclosed by the outer contour of the trophoblast is determined using Green's formula.

[0048] The present application provides a device for analyzing characteristics of blastocyst-stage embryos, the device comprising:

[0049] A contour extraction module is used to extract the contour of the trophoblast mask obtained by segmenting the blastocyst stage embryo image to obtain the trophoblast outer contour and the trophoblast inner contour;

[0050] A center point acquisition module, used to acquire the center point of the outer contour of the trophoblast;

[0051] An outer contour sampling module, used for sampling the outer contour of the trophoblast multiple times to obtain an outer contour sampling point set;

[0052] An inner contour sampling module, used to obtain the connection lines between each outer contour sampling point in the outer contour sampling point set and the center point, and obtain the inner contour sampling point set based on the intersection points between each of the connection lines and the inner contour of the trophoblast;

[0053] A distance determination module, for obtaining the distances between the outer contour sampling points and the inner contour sampling points corresponding to each sampling order based on the distances between the outer contour sampling points and the inner contour sampling points in the same sampling order;

[0054] The mapping analysis module is used to map the distances of the outer and inner point pairs corresponding to each sampling order to a coordinate system to obtain a waveform, and obtain the estimated number of cells on the trophoblast according to the number of peaks in the waveform.

[0055] The present application provides an embryo analysis system, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the above method.

[0056] The present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to execute the above method.

[0057] The present application provides a computer program product, on which a computer program is stored, and the computer program is executed by a processor to perform the above method.

[0058] In the above-mentioned method, device, embryo analysis system, storage medium and computer program product for analyzing the characteristics of blastocyst stage embryos, contour extraction is performed on the trophoblast mask obtained by segmenting the blastocyst stage embryo image, thereby obtaining the trophoblast outer contour and the trophoblast inner contour, then the center point of the trophoblast outer contour is obtained, and the trophoblast outer contour is sampled multiple times to obtain an outer contour sampling point set, and the connecting lines between each outer contour sampling point and the center point in the outer contour sampling point set are obtained, and the inner contour sampling point set is obtained based on the intersection points between each connecting line and the trophoblast inner contour; based on the same sampling times, the inner contour sampling point set is obtained. The distance between the outer contour sampling points and the inner contour sampling points of each sampling order is obtained, and the outer-inner point pair distance corresponding to each sampling order is mapped to the coordinate system, thereby realizing the spatial transformation of the trophoblast from a ring structure to a band structure, so that the number of cells on the trophoblast can be estimated according to the number of peaks in the obtained waveform, and the estimated number of cells is obtained. Compared with the end-to-end AI model, the above method provided by the present application has a transparent analysis process and interpretable feature parameters, which is more in line with the embryologist's evaluation process of the embryo. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A schematic diagram of a process for analyzing characteristics of blastocyst-stage embryos in one embodiment;

[0060] Figure 2 is a feature map of semantic segmentation in one embodiment;

[0061] FIG3( a ) is a schematic diagram of sampling results of trophoblast profile in one embodiment;

[0062] FIG3( b ) is a waveform diagram obtained after mapping the distances between inner and outer point pairs corresponding to each sampling to a coordinate system in one embodiment;

[0063] FIG3( c ) is a schematic diagram showing the marking of a peak in a waveform in one embodiment;

[0064] Figure 4 Schematic diagram of the implementation process of the Douglas-Peucker algorithm in one embodiment;

[0065] Figure 5 A schematic diagram of the sampling results of the trophoblast based on the Douglas-Peucker algorithm in one embodiment;

[0066] Figure 6 is a schematic diagram of the corresponding relationship between wave crests and cells in one embodiment;

[0067] Figure 7 is a schematic diagram of the outline of the inner cell mass in one embodiment;

[0068] Figure 8 A schematic diagram of a process for analyzing characteristics of blastocyst-stage embryos in one embodiment;

[0069] Fig. 9 A schematic diagram of trophoblast segmentation results in one embodiment;

[0070] Fig.10 A schematic diagram of the characteristic analysis results of the inner cell mass in one embodiment;

[0071] Fig.11 A graph showing a change trend of the blastocyst cavity ratio and a graph showing a change trend of the blastocyst expansion area in one embodiment;

[0072] Fig.12 It is a structural block diagram of a device for analyzing characteristics of blastocyst stage embryos in one embodiment;

[0073] Fig.13 FIG. 4 is a diagram showing the internal structure of an embryo analysis system in one embodiment. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0075] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0076] The main method for evaluating the embryonic potential of blastocysts is based on the morphological dynamics method. This application proposes a method for analyzing the characteristics of blastocyst-stage embryos. Through AI technology and image recognition technology, the morphological parameters of the blastocyst are automatically calculated. Therefore, the blastocyst-stage embryos can be evaluated based on the relevant blastocyst scoring consensus.

[0077] Based on the Istanbul consensus, the evaluation of blastocysts can be carried out on the trophoblast layer. The cells on the trophoblast layer are small cells that begin to differentiate and expand and arrange along the inner wall of the zona pellucida after the embryo develops into a morula. The development of the trophoblast layer can be divided into the following three levels according to the refined Istanbul consensus:

[0078] Grade A standard: The number of cells distributed along the "equatorial plane" of the blastocyst is significantly greater than 10, and they are uniform in size. The cells on the bottom surface of the blastocyst are all clear in morphology, and most of them have visible nuclei.

[0079] Grade B standard: There are about 10 cells distributed along the "equatorial plane" of the blastocyst, and the size is uneven. The cell morphology on the ground surface of the blastocyst is clear, and the cell nucleus is partially visible.

[0080] Grade C standard: The number of cells distributed along the "equatorial plane" of the blastocyst is significantly less than 10, the size is obviously uneven, there are obvious debris residues between the trophoblast cells and the zona pellucida, and the cells on the bottom surface of the blastocyst are difficult to identify.

[0081] Therefore, the analysis of trophoblast characteristics mainly includes estimating the number of cells on the trophoblast and quantifying the distribution of cells on the trophoblast. In terms of estimating the number of cells on the trophoblast, the present application provides a method for mapping the relationship between the inner and outer contours of the trophoblast into a two-dimensional coordinate system by sampling. The implementation process is similar to "straightening and flattening" the irregular ring. Through the expression of this mapping relationship, the waveform analysis of the inner and outer contours of the trophoblast is performed, thereby realizing the estimation of the number of cells on the trophoblast.

[0082] Specifically, the number of cells on the trophoblast layer is estimated by Figure 1 The steps shown are:

[0083] Step S101 , performing contour extraction on the trophoblast mask obtained by segmenting the blastocyst stage embryo image to obtain the trophoblast outer contour and the trophoblast inner contour.

[0084] After obtaining the blastocyst stage embryo image, semantic segmentation can be performed on the blastocyst stage embryo image to obtain the trophoblast mask. In addition, the inner cell mass mask, the blastocyst cavity mask, the zona pellucida (ZP) mask, and the blastocyst body mask can also be obtained.

[0085] The semantic segmentation method can be built based on networks such as deeplabV3, Unet, UneXt, and FCN. More specifically, FCN (Fully Convolutional Networks) can be used to classify blastocyst embryo images at the pixel level, thereby solving the image segmentation problem at the semantic level; FCN can accept input images of any size, and use deconvolution to upsample the featuremap of the last convolutional layer to restore it to the same size as the input image, so that a prediction can be generated for each pixel while retaining the spatial information in the original input image. Finally, pixel-by-pixel classification is performed on the upsampled feature map. Figure 2 is the feature map of semantic segmentation, Figure 2 The five parts shown in the figure represent the five components of the blastocyst that need to be analyzed. Figure 2 (a) corresponds to the main body of the blastocyst. Figure 2 (b) corresponds to the blastocyst cavity. Figure 2 (c) corresponds to the inner cell mass. Figure 2(d) corresponds to the trophoblast, Figure 2 (e) corresponds to the zona pellucida.

[0086] In order to achieve better segmentation performance, we can use the FCN semantic segmentation network based on ResNet-101 and use the convolutional neural network of ResNet-101 for downsampling. The benefits of this are as follows: a. The number of network layers is deepened, which improves the segmentation accuracy of the network; b. More jump connections can be added in the middle of the network, so that the background semantic information of the image can be better combined for multi-scale segmentation; c. ResNet-101 has the advantages of fast convergence and reduced model data volume; d. ResNet-101 makes the model easier to train, which can prevent model degradation, gradient disappearance, and loss non-convergence.

[0087] Segmentation training uses data annotation tools to select the dataset of blastocyst stage (where obvious blastocyst cavity area appears) for labeling, mark the segmentation area, set the classification label, and the label is divided into five categories: 0: blastocyst body, 1: blastocyst cavity, 2: inner cell mass, 3: trophoblast, 4: zona pellucida. The labeled dataset is divided into training set and validation set at a ratio of 3:1, and the input raw data is planned and data enhanced, and then iterative training is performed to obtain the inference model, and then the masks of the feature areas are obtained according to the inference model: trophoblast mask (TE_mask), inner cell mass mask (ICM_mask) and blastocyst cavity mask (Cavity_mask).

[0088] The trophoblast outline TE_Polygon is extracted through the trophoblast mask and the outline is split into inner and outer closed outlines: the trophoblast outer outline TE_Polygon_out and the trophoblast inner outline TE_Polygon_in.

[0089] Among them, the contour extraction method can adopt the contour approximation method, which is implemented based on the findContours function in opencv. The detection mode is set to RETR_TREE, all contours are retrieved and the complete hierarchy of nested contours is reconstructed. The enumerator selection of the contour approximation method is: CHAIN_APPROX_SIMPLE.

[0090] Step S102, obtaining the center point of the outer contour of the trophoblast, and sampling the outer contour of the trophoblast multiple times to obtain an outer contour sampling point set.

[0091] Step S103, obtaining the connecting lines between each outer contour sampling point and the center point in the outer contour sampling point set, and obtaining the inner contour sampling point set based on the intersection points between each connecting line and the inner contour of the trophoblast.

[0092] Step S104, based on the distances between the outer contour sampling points and the inner contour sampling points of the same sampling order, obtain the outer-inner point pair distances corresponding to each sampling order.

[0093] Step S105, mapping the distances between the outer and inner point pairs corresponding to each sampling order to a coordinate system to obtain a waveform, and obtaining an estimated number of cells on the trophoblast according to the number of peaks in the waveform.

[0094] Among them, each time the outer contour of the trophoblast is sampled once, an outer contour sampling point can be obtained. After completing one sampling, the next sampling is performed according to the preset sampling step length to obtain the next outer contour sampling point. Therefore, after the outer contour of the trophoblast is sampled multiple times, the outer contour sampling points corresponding to each sampling can be obtained to form an outer contour sampling point set, which can be defined as C n , n represents the outer contour.

[0095] For example, after obtaining the outer contour sampling point set C n After that, the outer contour sampling points (C nx , C ny ), based on (C nx , C ny ) and the center point O of the outer contour of the trophoblast T The line between the two is used to obtain the linear equation L corresponding to the first sampling. n .

[0096] Next, calculate the L corresponding to the first sampling n The intersection with the inner contour of the trophoblast (C mx , C my ), the intersection point (C mx , C my ) is the inner contour sampling point corresponding to the first sampling, and (C mx , C my ) is stored in the inner contour sampling point set C m In the figure, m represents the inner contour.

[0097] Similarly, the inner contour sampling points corresponding to other sub-samplings can be obtained in the above manner.

[0098] The sampling results of the outer contour and the inner contour of the trophoblast are shown in Figure 3(a). After the sampling is completed, the distance k between the outer contour sampling point corresponding to the first sampling and the inner contour sampling point corresponding to the first sampling can be calculated. nm [1], we can calculate the distance k between the outer contour sampling point corresponding to the i-th sampling and the inner contour sampling point corresponding to the i-th sampling. nm [i], thus, we can get the distances {k nm [1]...k nm[i]}.

[0099] Next, the distances of the outer and inner points from the first sampling to the i-th sampling are calculated in such a way that the sampling order corresponds to the horizontal axis of the coordinate system and the distances of the outer and inner points correspond to the vertical axis of the coordinate system. nm [1]...k nm [i]}, mapped to a two-dimensional coordinate system, the discrete waveform shown in Figure 3(b) is obtained. Since the number of peaks of the waveform is approximately equal to the number of cells in the trophoblast, after obtaining the waveform, the peaks in the waveform can be marked, as shown in Figure 3(c). According to the number of peaks, the number of cells on the trophoblast is estimated to obtain the estimated number of cells. The horizontal axis in Figures 3(b) and 3(c) represents the sampling step size, and also represents the circumference of the outer circle of the trophoblast, in pixels. The vertical axis represents the distance between the outer and inner points.

[0100] In the above-mentioned method for analyzing the characteristics of blastocyst stage embryos, the contour of the trophoblast mask obtained by segmenting the blastocyst stage embryo image is extracted to obtain the outer contour and the inner contour of the trophoblast. Then, the center point of the outer contour of the trophoblast is obtained, and the outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set, and the lines between each outer contour sampling point and the center point in the outer contour sampling point set are obtained. Based on the intersection between each line and the inner contour of the trophoblast, the inner contour sampling point set is obtained; based on the distance between the outer contour sampling point and the inner contour sampling point of the same sampling order, the outer and inner point pair distances corresponding to each sampling order are obtained, and the outer and inner point pair distances corresponding to each sampling order are mapped to the coordinate system, thereby realizing the spatial transformation of the trophoblast from a ring structure to a band structure, so that the number of cells on the trophoblast can be estimated according to the number of peaks in the obtained waveform, and the estimated number of cells is obtained. Compared with the end-to-end AI model, the above-mentioned method provided by the present application has a transparent analysis process and interpretable feature parameters, which is more in line with the embryologist's evaluation process of the embryo.

[0101] In one embodiment, obtaining the center point of the outer contour of the trophoblast layer may specifically include the following steps: obtaining the zero-order spatial moment, the first-order spatial moment about the horizontal axis, and the first-order spatial moment about the vertical axis of the image area enclosed by the outer contour of the trophoblast layer; determining the center point of the outer contour of the trophoblast layer in the horizontal and vertical directions according to the ratio of the first-order spatial moment about the horizontal axis to the zero-order spatial moment, and the ratio of the first-order spatial moment about the vertical axis to the zero-order spatial moment.

[0102] In this embodiment, the image area surrounded by the outer contour of the trophoblast can be determined, and the zero-order spatial moment m of the image area can be calculated. 00 , the first-order spatial moment m about the horizontal axis 01 and the first-order spatial moment m about the vertical axis 10 .

[0103] The calculation formula of spatial moment is: Among them, array(x, y) is the pixel value at pixel (x, y). When i = j = 0, it corresponds to the zero-order spatial moment; when i = 0 and j = 1, it corresponds to the first-order spatial moment about the horizontal axis; when i = 1 and j = 0, it corresponds to the first-order spatial moment about the vertical axis.

[0104] For a binarized image, the zero-order spatial moment m 00 is the area of the image region enclosed by the outer contour of the trophoblast.

[0105] Next, the ratio of the first-order spatial moment m 01 about the horizontal axis to the zero-order spatial moment m 00 can be obtained, and the ratio of the first-order spatial moment m 10 about the vertical axis to the zero-order spatial moment m 00 is obtained, so as to obtain the coordinates T of the center point O of the outer contour of the trophoblast.

[0106] In the above embodiment, from the perspective of image spatial moments, according to the ratio of the first-order spatial moment about the horizontal axis to the zero-order spatial moment, and the ratio of the first-order spatial moment about the vertical axis to the zero-order spatial moment, the center point of the outer contour of the trophoblast is accurately determined in the horizontal axis direction and the vertical axis direction.

[0107] In one embodiment, the outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set, including: sampling the outer contour of the trophoblast multiple times in a uniform sampling manner according to the azimuth angle in 360°, to obtain an outer contour sampling point set; or, sampling the outer contour of the trophoblast multiple times according to the sampling method of the Douglas-Peucker algorithm, to obtain an outer contour sampling point set.

[0108] The basic idea of the Douglas-Peucker algorithm is: Virtually connect a straight line between the start and end points of each curve, find the distances from all points to the straight line, and find the maximum distance value dmax. Compare dmax with the tolerance D: If dmax < D, all the intermediate points on this curve are discarded. If dmax ≥ D, retain the coordinate points corresponding to dmax, and take this point as the boundary to divide the curve into two parts, and repeat this method for these two parts. The tolerance D = L per ×δ; where L per is the contour perimeter, δ is the sampling coefficient, the sampling coefficient is set as a constant, the smaller the coefficient, the denser the sampling points, and the larger the coefficient, the sparser the sampling points.

[0109] Specific implementation process:

[0110] ① Virtually connect a straight line between the two end points of the curve, and find the distances from the remaining points to this straight line, such as Figure 4 (a) in the.

[0111] ② Select the largest one and compare it with the threshold. If it is greater than the threshold, the point with the largest distance from the straight line is retained, otherwise all points between the two end points of the straight line are discarded. Figure 4 In (b), point 4 is retained.

[0112] ③ According to the retained points, the known curve is divided into two parts, and steps ① and ② are repeated, and the operation is iterative, that is, the one with the largest distance is still selected to compare with the threshold, and then selected in turn until there are no points to be discarded, and finally the coordinates of the curve points that meet the given accuracy limit are obtained, such as Figure 4 In (c) and (d), the 6th and 7th points are retained respectively, and the other points are discarded, thus completing the simplification of the contour.

[0113] Therefore, according to the sampling method of the Douglas-Peucker algorithm, the outer contour of the trophoblast is sampled multiple times to obtain the outer contour sampling point set C n .

[0114] In the above embodiment, when downsampling the outer contour of the trophoblast, the Douglas-Peucker algorithm can be selected. Compared with the method of using azimuth moving traversal to achieve downsampling of the outer contour of the trophoblast, the technology implemented by the Douglas-Peucker algorithm is simpler, and the sparse and dense sampling can be flexibly adjusted by setting the sampling coefficient δ.

[0115] Figure 5 In order to sample the trophoblast based on the Douglas-Peucker algorithm, when the trophoblast mask is missing, the gap can be filled by ellipse fitting, and then the trophoblast mask can be analyzed and the number of cells on the trophoblast can be estimated.

[0116] In one embodiment, the estimated number of cells on the trophoblast is obtained based on the number of peaks in the waveform, which specifically includes the following steps: obtaining the total number of peaks in the waveform and the width of each peak; subtracting the number of peaks whose widths do not fall within a preset interval from the total number of peaks; the preset interval is determined based on the number of pixel units occupied by the cells; and using the subtraction result as the estimated number of cells on the trophoblast.

[0117] For example, the width of a cell is generally between [15, 60] pixel units. After obtaining the total number of peaks in the waveform, the number of peaks whose width does not fall within [15, 60] can be subtracted from the total number of peaks to obtain a subtraction result, which is used as the estimated number of cells, that is, the estimated number of cells = the total number of peaks - the number of peaks whose width is outside [15, 60] pixel units.

[0118] In the above embodiment, referring to Figure 6It can be seen that the number of peaks of the waveform is approximately equal to the number of cells on the trophoblast, and the width of the peaks is approximately equal to the width of the cells. Therefore, in order to further improve the accuracy of the estimation results, this embodiment uses the width range of the cells as a screening criterion based on the number of peaks to filter out erroneously counted peaks.

[0119] In one embodiment, after obtaining the estimated number of cells on the trophoblast, the method provided in the present application may further include the following steps: quantifying the distribution of cells on the trophoblast based on the circumference of the outer contour of the trophoblast and the estimated number of cells to obtain a quantitative distribution result.

[0120] In this embodiment, after obtaining the circumference of the outer contour of the trophoblast and the estimated number of cells, the estimated number of cells can be divided by the circumference of the outer contour of the trophoblast to obtain a ratio: estimated number of cells / circumference of the outer contour of the trophoblast, thereby quantifying the distribution of cells on the trophoblast; the larger the above ratio, the denser the distribution of cells on the trophoblast, thereby obtaining a quantitative distribution result.

[0121] In the above embodiment, the distribution of cells on the trophoblast is quantified according to the perimeter of the outer contour of the trophoblast and the estimated number of cells, so that the distribution of cells on the trophoblast can be known more intuitively.

[0122] Based on the Istanbul consensus, the evaluation of the blastocyst can also be carried out on the inner cell mass. After the embryo develops into a morula, the morula develops further and the cells begin to differentiate. The cells that gather at one end of the embryo and are larger in size are called the inner cell mass. The cells of the inner cell mass are embryonic stem cells, which are undifferentiated cells with developmental omnipotence and will develop into various tissues of the fetus in the future.

[0123] According to the consensus proposed by ASEBIR (Spanish Association for Reproductive Research), new criteria for the developmental quality of the inner cell mass have been made:

[0124] Grade A: The surface area of ​​the inner cell mass is 1900 to 3800um 2 The inner cell mass is tight.

[0125] Grade B: The surface area of ​​the inner cell mass is 1900 to 3800um 2 In between, the inner cell mass is loose.

[0126] Grade C: The surface area of ​​the inner cell mass is less than 1900um 2 .

[0127] Based on the above consensus description of the inner cell mass grade, the characteristic analysis of the inner cell mass is mainly quantitatively described from the surface area of ​​the inner cell mass, the degree of densification of the inner cell mass, etc. In addition, considering the characteristics of the inner cell mass clustering, the roundness characteristics of the inner cell mass can also be used as a reference evaluation standard.

[0128] Based on this, the method provided in the present application may further include the following steps: extracting the contour of the inner cell mass mask in the blastocyst stage embryo image to obtain the inner cell mass contour (such as Figure 7 Based on the outline of the inner cell mass, the perimeter and area of ​​the inner cell mass are obtained; based on the perimeter and area of ​​the inner cell mass, the roundness of the inner cell mass is obtained.

[0129] When the circularity of the inner cell mass is obtained based on the perimeter and area of ​​the inner cell mass, the formula Find the circularity of the inner cell mass, L i The perimeter of the inner cell mass is S i is the inner cell mass area, R i is the degree of inner cell roundness.

[0130] In one embodiment, the characteristic analysis of the inner cell mass may include analyzing the movement trajectory of the inner cell mass. In this case, the method provided in the present application also includes the following steps: collecting images of the embryo during the blastocyst stage development, obtaining images of the blastocyst stage embryo at multiple imaging moments, to obtain the contour of the inner cell mass corresponding to each imaging moment; according to the contour of the inner cell mass corresponding to each imaging moment, obtaining the center point position of the inner cell mass at each imaging moment; mapping the center point position of the inner cell mass at each imaging moment to the coordinate system, to obtain the movement trajectory of the inner cell mass during the blastocyst stage development.

[0131] In this embodiment, images of embryos during the blastocyst stage development process can be collected to obtain blastocyst stage embryo images at multiple imaging moments, and then semantic segmentation is performed on each blastocyst stage embryo image to obtain an inner cell mass mask corresponding to each blastocyst stage embryo image, and then contour extraction is performed on each inner cell mass mask to obtain the inner cell mass contour corresponding to each imaging moment. The contour extraction method can be consistent with the above-mentioned trophoblast contour extraction method.

[0132] Next, the center point position of the inner cell mass at each imaging moment can be obtained based on the inner cell mass contour corresponding to each imaging moment in accordance with the above-mentioned method of calculating the center point corresponding to the outer contour of the trophoblast. Each center point position can be determined based on the horizontal position and the vertical position of the center point on the image. Therefore, the horizontal position of the center point in the image can be corresponded to the x-axis of the coordinate system, and the vertical position of the center point in the image can be corresponded to the y-axis of the coordinate system. According to this correspondence, the center point position of the inner cell mass at each imaging moment is mapped to the same two-dimensional coordinate system. After each center point position is mapped to the two-dimensional coordinate system, the corresponding imaging moment can be marked, thereby obtaining the movement trajectory of the inner cell mass during the blastocyst stage development.

[0133] In one embodiment, the method provided in the present application also includes: collecting images of the embryo during the blastocyst stage development process to obtain blastocyst stage embryo images at multiple imaging moments; extracting the corresponding blastocyst cavity mask and trophoblast outer contour from the blastocyst stage embryo images corresponding to the same imaging moment; obtaining the blastocyst cavity ratio corresponding to each of the multiple imaging moments based on the image area enclosed by the blastocyst cavity mask and trophoblast outer contour corresponding to the same imaging moment; obtaining the blastocyst cavity change trend based on the blastocyst cavity ratio corresponding to each of the multiple imaging moments.

[0134] In this embodiment, images of embryos during the blastocyst stage development process can be collected to obtain multiple blastocyst stage embryo images at the imaging moments, and then semantic segmentation is performed on each blastocyst stage embryo image, and the blastocyst cavity mask and trophoblast outer contour at the corresponding imaging moment are extracted from each blastocyst stage embryo image. Then, based on the image area area enclosed by the blastocyst cavity mask and trophoblast outer contour corresponding to the same imaging moment, the blastocyst cavity ratio corresponding to each of the multiple imaging moments is obtained; based on the blastocyst cavity ratio corresponding to each of the multiple imaging moments, the blastocyst cavity change trend is obtained. Among them, the contour extraction method can be consistent with the above-mentioned trophoblast contour extraction method.

[0135] In order to segment out a more complete trophoblast and inner cell mass structure, we can select embryo images of the third stage of blastocyst (which can be recorded as blastocyst stage III) or the fourth stage of blastocyst (blastocyst stage IV), perform trophoblast mask extraction and inner cell mass mask extraction, analyze trophoblast-related features and inner cell mass-related features, and further analyze the developmental potential of the blastocyst.

[0136] According to Gardner's blastocyst staging principles:

[0137] Blastocyst stage I (can be recorded as blastocyst stage I): early blastocysts have chambers, and the volume of the blastocyst cavity is less than 1 / 2 of the total volume of the embryo;

[0138] Blastocyst stage II (can be recorded as blastocyst stage II): the volume of the blastocyst cavity is greater than or equal to 1 / 2 of the embryo volume;

[0139] Blastocyst stage III: The blastocyst cavity completely occupies the total volume of the embryo;

[0140] Blastocyst stage IV: The blastocyst expands, the blastocyst cavity completely fills the embryo, the total volume of the embryo increases, and the zona pellucida becomes thinner.

[0141] Based on this, the method provided in the present application also includes: obtaining a blastocyst cavity mask extracted from a blastocyst stage embryo image; obtaining the blastocyst cavity ratio based on the area of ​​the blastocyst cavity mask and the area of ​​the image area enclosed by the trophoblast outer contour; determining whether the embryo is in the blastocyst stage III or blastocyst stage IV based on the blastocyst cavity ratio; if the embryo is in the blastocyst stage III or blastocyst stage IV, executing the steps after contour extraction of the trophoblast mask in the blastocyst stage embryo image, that is, obtaining the center point of the trophoblast outer contour, and sampling the trophoblast outer contour multiple times to obtain an outer contour sampling point set; or, if the embryo is in the blastocyst stage III or blastocyst stage IV, performing contour extraction on the inner cell mass mask in the blastocyst stage embryo image.

[0142] Exemplarily, the blastocyst cavity ratio Among them, Cavity_mask_area is the area of ​​the blastocyst cavity mask, and Blastaea_mask_area is the area of ​​the image region enclosed by the outer contour of the trophoblast.

[0143] Furthermore, when obtaining the area Blastaea_mask_area of ​​the region enclosed by the outer contour of the trophoblast, the following steps may be specifically included: determining the area of ​​the image region enclosed by the outer contour of the trophoblast according to the number of pixels within the outer contour of the trophoblast; or, determining the area of ​​the image region enclosed by the outer contour of the trophoblast using Green's formula.

[0144] Among them, Green's formula is specifically: suppose the closed area D is surrounded by a piecewise smooth curve L, and the functions P(x, y) and Q(x, y) have first-order continuous partial derivatives on D, then:

[0145] Where L is the positive boundary curve of D.

[0146] In order to better understand the above method, an embodiment of the method for analyzing the characteristics of blastocyst embryos in the present application is described in detail below. The method provided in the present application can be applied to the software analysis system of the Time-lapse (time-lapse imaging) embryo incubator, and the blastocyst data on the 5th or 6th day provided by Time-lapse can be automatically analyzed, and the characteristics of the trophoblast and inner cell mass of the blastocyst can be extracted and quantified, and the developmental potential of the blastocyst can be automatically scored based on the quantitative data and the scoring consensus.

[0147] This embodiment performs blastocyst feature analysis based on the data provided by Time-lapse, including the following six parts: Figure 8 The process structure is shown as follows:

[0148] (1) Semantic segmentation: extract masks of regions such as the trophoblast, inner cell mass, blastocyst cavity, and blastocyst body.

[0149] Specifically, the FCN semantic segmentation network based on ResNet-101 can be used to perform semantic segmentation on the blastocyst stage embryo image to obtain the trophoblast mask, inner cell mass mask, blastocyst cavity mask, zona pellucida mask, and blastocyst body mask.

[0150] (2) Blastocyst staging: The developmental period of the blastocyst is inferred based on the change in the area ratio of the blastocyst cavity to the blastocyst body. Specifically:

[0151] Step S1, extracting the contour of the trophoblast mask to obtain the outer contour of the trophoblast;

[0152] Step S2, calculating the area of ​​the blastocyst cavity mask and the area of ​​the image region enclosed by the outer contour of the trophoblast;

[0153] There are two ways to calculate the area. One is to calculate the area of ​​the connected domain based on the number of pixels within the edge, and the other is to use Green's formula to calculate the area.

[0154] Step S3, obtaining the blastocyst cavity ratio according to the area of ​​the blastocyst cavity mask and the area of ​​the image region enclosed by the outer contour of the trophoblast;

[0155] Step S4, based on the blastocyst cavity ratio, determine whether the embryo is in the blastocyst stage III or blastocyst stage IV.

[0156] (3) Trophoblast analysis: Select embryo images of blastocyst stage III and blastocyst stage IV for trophoblast segmentation and extract trophoblast-related features, such as Fig. 9 shown.

[0157] Trophoblast characterization can include: ① estimating the number of cells on the trophoblast and ② quantifying the distribution of cells.

[0158] ① Estimate the number of cells on the trophoblast, including the following steps:

[0159] Step S5, obtaining the center point of the outer contour of the trophoblast;

[0160] Specifically, the zero-order spatial moment of the image area enclosed by the outer contour of the trophoblast layer, the first-order spatial moment about the horizontal axis, and the first-order spatial moment about the vertical axis are obtained; according to the ratio of the first-order spatial moment about the horizontal axis to the zero-order spatial moment, and the ratio of the first-order spatial moment about the vertical axis to the zero-order spatial moment, the center point of the outer contour of the trophoblast layer is determined in the horizontal and vertical directions.

[0161] Step S6, sampling the outer contour of the trophoblast multiple times to obtain an outer contour sampling point set;

[0162] Specifically, the sampling method may be: uniform sampling in 360° according to the azimuth angle, or sampling based on the Douglas-Peucker algorithm.

[0163] Step S7, obtaining the connecting lines between each outer contour sampling point and the center point in the outer contour sampling point set, and obtaining the inner contour sampling point set based on the intersection points between each connecting line and the inner contour of the trophoblast;

[0164] Step S8, based on the distances between the outer contour sampling points and the inner contour sampling points in the same sampling order, obtaining the outer and inner point pair distances corresponding to each sampling order;

[0165] Step S9, mapping the distances of the outer and inner point pairs corresponding to each sampling order to a coordinate system to obtain a waveform;

[0166] Step S10, obtaining the total number of peaks in the waveform and the width of each peak; from the total number of peaks, subtracting the number of peaks whose widths do not fall within a preset interval; the preset interval is determined based on the number of pixel units occupied by cells; and using the subtraction result as an estimated number of cells on the trophoblast.

[0167] ② Quantify the distribution of cells, including the following steps:

[0168] The distribution of cells on the trophoblast is quantified according to the perimeter of the outer contour of the trophoblast and the estimated number of cells to obtain a quantitative distribution result.

[0169] (4) Inner cell mass analysis: Fig.10 As shown, embryo images of blastocyst stage III and blastocyst stage IV are selected to segment the inner cell mass, extract the relevant features of the inner cell mass, and record the movement trajectory of the inner cell mass.

[0170] The analysis of inner cell mass characteristics may include: ① calculating the inner cell mass roundness and ② calculating the movement trajectory of the inner cell mass;

[0171] ① Calculate the inner cell roundness, including the following steps:

[0172] Step S11, extracting the contour of the inner cell mass mask in the blastocyst stage embryo image to obtain the inner cell mass contour;

[0173] Step S12, obtaining the perimeter and area of ​​the inner cell mass based on the inner cell mass contour;

[0174] Step S13, obtaining the roundness of the inner cell mass according to the perimeter and area of ​​the inner cell mass.

[0175] ② Calculate the movement trajectory of the inner cell mass, including the following steps:

[0176] Step S14, collecting images of the embryo during the blastocyst stage development process, obtaining images of the blastocyst stage embryo at multiple imaging moments, so as to obtain the inner cell mass contour corresponding to each imaging moment;

[0177] Step S15, obtaining the center point position of the inner cell mass at each imaging moment according to the inner cell mass contour corresponding to each imaging moment;

[0178] Step S16, mapping the center point position of the inner cell mass at each imaging moment into the coordinate system, and obtaining the movement trajectory of the inner cell mass during the blastocyst stage development.

[0179] (5) Blastocyst expansion time series analysis: Record the trend of blastocyst expansion over time, and analyze the abnormal development of blastocysts from the time series graph. For example, time series analysis can be used to determine whether the cavity shrinks during the expansion process. Fig.11 As shown, the horizontal axis represents the culture time of the embryo from the moment of fertilization, in hours.

[0180] (6) The inner cell mass and trophoblast cell layer were graded according to the Gardner evaluation system.

[0181] The above-mentioned process of characterizing blastocysts includes the following key technologies:

[0182] 1. In the process of semantic segmentation of blastocyst features, since blastocyst images have complex background textures and are interfered by artifacts and other imaging factors, it is difficult to identify and mark targets. In order to obtain good segmentation accuracy, multiple semantic segmentation methods were tested, including segmentation networks such as deeplabV3, Unet, UneXt and FCN. Table 1 shows the model performance segmentation test results of the three methods of Unet, FCN-Resnet101 and deeplabV3. FCN-Resnet101 is more suitable for completing the segmentation task of blastocyst features, balancing the requirements for detail and classification accuracy of feature extraction.

[0183] 2. The Douglas-Peucker algorithm is selected for contour downsampling. Compared with the method of using azimuth moving traversal to achieve contour sampling, the technology implemented by the Douglas-Peucker algorithm is simpler, and the sparse and dense sampling can be flexibly adjusted by setting the sampling coefficient δ. Figure 5 This is the result of trophoblast sampling based on the Douglas-Peucker algorithm. When the trophoblast mask is missing, the "gap filling" method can be used to analyze the trophoblast mask and estimate the number of trophoblast cells.

[0184] 3. The mapping technology of the trophoblast annulus into a two-dimensional coordinate system realizes the spatial transformation from the annular structure to the band structure, so that the number of trophoblast cells can be estimated by analyzing the waveform of the mapping relationship.

[0185]

[0186] Table 1

[0187] Based on the Istanbul Consensus and ASEBIR Consensus, the blastocyst embryo data obtained by the Time-lapse imaging system is automatically analyzed and rated for developmental potential through AI technology and image analysis algorithms. Compared with the current blastocyst potential evaluation algorithms and manual evaluation methods, this application can be more intelligent, transparent and objective from the perspective of technical principles and implementation processes. The method provided in this application can better assist embryologists in convenient and rapid embryo selection, can more comprehensively extract relevant characteristic factors of embryo development, and provide good assistance to embryologists' data backtracking and embryo research. The advantages of this application are specifically:

[0188] 1. Feature extraction of the trophoblast, inner cell mass, and blastocyst cavity of the blastocyst can more objectively and comprehensively analyze the developmental potential of the blastocyst;

[0189] 2. Calculation and analysis of multiple characteristic indicators of trophoblast, inner cell mass, etc. can assist embryologists in analyzing potential indicators of blastocyst development potential.

[0190] 3. The characteristic analysis standards are based on expert consensus standards, and the analysis process is closer to the analysis principles of embryologists.

[0191] 4. Continuously track the expansion status of the blastocyst and analyze the development process of the blastocyst in a time-series manner.

[0192] 5. Provide a method of extracting target features through AI technology and combining image analysis technology to quantitatively analyze target features, so that AI technology can analyze embryos in a transparent, explainable and visible way. The method provided in this embodiment can be used as part of an auxiliary decision-making tool. The final result achieved is not only to provide embryologists with a level of prediction of blastocyst potential, but also to pay more attention to the feature analysis details in the process. The quantitative results of the open features can enable embryologists to select embryos in the blastocyst cavity more objectively and accurately.

[0193] 6. Since the formation and development of blastocyst is a complex dynamic process, and there are various vacuoles, fragments, cleavage cells and other substances inside the blastocyst, the traditional analysis method of trophoblast characteristics can only be based on artificial fuzzy judgment. This embodiment proposes a method for estimating and counting the number of trophoblast cells, thereby achieving the purpose of using the number of trophoblast cells as a quantitative feature for potential assessment.

[0194] 7. When analyzing the characteristics of the inner cell mass, not only can the morphological characteristics and roundness of the inner cell mass be obtained, but also the movement trajectory of the inner cell mass can be tracked and the degree of compaction of the inner cell mass can be quantified, which enriches the evaluation indicators of the inner cell mass and assists embryologists in observing and discovering more evaluation indicators for the direction of embryonic development.

[0195] 8. Since the development process of the embryo is a time-series process, this embodiment associates the proportion of the blastocyst cavity, the expansion area of ​​the blastocyst with the fertilization time of the embryo, and obtains the changing trends of the proportion of the blastocyst cavity and the expansion area of ​​the blastocyst with the fertilization time during embryo development, which can assist embryologists in analyzing the correlation between the degree of blastocyst expansion and the embryo's development potential.

[0196] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0197] In one embodiment, Fig.12 As shown, a device for analyzing characteristics of blastocyst stage embryos is provided, comprising:

[0198] Contour extraction module 1201, used to extract contours of the trophoblast mask obtained by segmenting the blastocyst stage embryo image, to obtain the trophoblast outer contour and the trophoblast inner contour;

[0199] A center point acquisition module 1202 is used to acquire the center point of the outer contour of the trophoblast;

[0200] An outer contour sampling module 1203 is used to perform multiple sampling on the outer contour of the trophoblast to obtain an outer contour sampling point set;

[0201] The inner contour sampling module 1204 is used to obtain the connection lines between each outer contour sampling point in the outer contour sampling point set and the center point, and obtain the inner contour sampling point set based on the intersection points between each of the connection lines and the inner contour of the trophoblast;

[0202] The distance determination module 1205 is used to obtain the distances between the outer contour sampling points and the inner contour sampling points corresponding to each sampling order based on the distances between the outer contour sampling points and the inner contour sampling points of the same sampling order;

[0203] The mapping analysis module 1206 is used to map the distances between the outer and inner point pairs corresponding to each sampling order to a coordinate system to obtain a waveform, and obtain an estimated number of cells on the trophoblast according to the number of peaks in the waveform.

[0204] In one embodiment, the center point acquisition module 1202 is also used to obtain the zero-order spatial moment of the image area enclosed by the outer contour of the trophoblast layer, the first-order spatial moment about the horizontal axis, and the first-order spatial moment about the vertical axis; according to the ratio of the first-order spatial moment about the horizontal axis to the zero-order spatial moment, and the ratio of the first-order spatial moment about the vertical axis to the zero-order spatial moment, the center point of the outer contour of the trophoblast layer is determined in the horizontal and vertical directions.

[0205] In one embodiment, the outer contour sampling module 1203 is also used to sample the outer contour of the trophoblast layer multiple times in a manner of uniformly sampling in 360° azimuth angle to obtain an outer contour sampling point set; or, sample the outer contour of the trophoblast layer multiple times in a sampling manner of the Douglas-Peucker algorithm to obtain an outer contour sampling point set.

[0206] In one embodiment, the mapping analysis module 1206 is also used to obtain the total number of peaks in the waveform and the width of each peak; from the total number of peaks, subtract the number of peaks whose widths do not fall within a preset interval; the preset interval is determined based on the number of pixel units occupied by cells; and the subtraction result is used as an estimated number of cells on the trophoblast.

[0207] In one embodiment, the device further comprises a distribution quantification module for quantifying the distribution of the cells on the trophoblast according to the perimeter of the outer contour of the trophoblast and the estimated number of the cells to obtain a quantitative distribution result.

[0208] In one embodiment, the device also includes an inner cell mass analysis module, which is used to extract the contour of the inner cell mass mask in the blastocyst stage embryo image to obtain the inner cell mass contour; based on the inner cell mass contour, obtain the circumference and area of ​​the inner cell mass; based on the circumference and area of ​​the inner cell mass, obtain the roundness of the inner cell mass.

[0209] In one embodiment, the inner cell mass analysis module is also used to collect images of the embryo during the blastocyst stage development, obtain images of the blastocyst stage embryo at multiple imaging moments, and obtain the inner cell mass contour corresponding to each imaging moment; according to the inner cell mass contour corresponding to each imaging moment, obtain the center point position of the inner cell mass at each imaging moment; map the center point position of the inner cell mass at each imaging moment to the coordinate system, and obtain the movement trajectory of the inner cell mass during the blastocyst stage development.

[0210] In one embodiment, the device also includes a cavity change trend acquisition module, which is used to collect images of the embryo during the blastocyst stage development process, and obtain blastocyst stage embryo images at multiple imaging moments; extract the corresponding blastocyst cavity mask and trophoblast outer contour from the blastocyst stage embryo image corresponding to the same imaging moment; obtain the blastocyst cavity ratio corresponding to each of the multiple imaging moments based on the image area enclosed by the blastocyst cavity mask and trophoblast outer contour corresponding to the same imaging moment; obtain the blastocyst cavity change trend based on the blastocyst cavity ratio corresponding to each of the multiple imaging moments.

[0211] In one embodiment, the device further comprises a blastocyst staging module, which is used to obtain a blastocyst cavity mask extracted from the blastocyst stage embryo image; obtain a blastocyst cavity ratio according to the area of ​​the blastocyst cavity mask and the area of ​​the image region enclosed by the outer contour of the trophoblast; and determine whether the embryo is in the blastocyst stage III or blastocyst stage IV based on the blastocyst cavity ratio;

[0212] If the embryo is in the blastocyst stage III or blastocyst stage IV, the step of performing contour extraction on the trophoblast mask in the blastocyst stage embryo image in the contour extraction module is executed, or the step of performing contour extraction on the inner cell mass mask in the blastocyst stage embryo image in the inner cell mass analysis module is executed.

[0213] In one embodiment, the device also includes an area calculation module for determining the area of ​​the image region enclosed by the outer contour of the trophoblast based on the number of pixels within the outer contour of the trophoblast; or, using Green's formula, determining the area of ​​the image region enclosed by the outer contour of the trophoblast.

[0214] The specific limitation of the device for blastocyst stage embryo characteristic analysis can be found in the limitation of the method for blastocyst stage embryo characteristic analysis above, which will not be repeated here. Each module in the device for blastocyst stage embryo characteristic analysis can be implemented in whole or in part by software, hardware and a combination thereof. Each of the modules can be embedded in or independent of a processor in an embryo analysis system in hardware form, or can be stored in a memory in an embryo analysis system in software form, so that the processor calls and executes the operations corresponding to the above modules.

[0215] In one embodiment, an embryo analysis system is provided, whose internal structure diagram can be as follows: Fig.13 As shown. The embryo analysis system includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the embryo analysis system is used to provide computing and control capabilities. The memory of the embryo analysis system includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the embryo analysis system is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. The embryo analysis system also includes an input and output interface, which is a connection circuit for exchanging information between the processor and an external device. They are connected to the processor through a bus, referred to as an I / O interface. When the computer program is executed by the processor, a method for analyzing the characteristics of blastocyst embryos is implemented. The display screen of the embryo analysis system can be a liquid crystal display screen or an electronic ink display screen, and the input device of the embryo analysis system can be a touch layer covered on the display screen, or a key, trackball or touchpad set on the shell of the embryo analysis system, or an external keyboard, touchpad or mouse.

[0216] Those skilled in the art will understand that Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the embryo analysis system to which the present application scheme is applied. The specific embryo analysis system may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0217] In one embodiment, an embryo analysis system is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps in the above-mentioned various method embodiments when executing the computer program.

[0218] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0219] In one embodiment, a computer program product is provided, on which a computer program is stored, and the computer program is used by a processor to execute the steps in the above-mentioned various method embodiments.

[0220] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0221] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the above-mentioned computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0222] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0223] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for analyzing characteristics of blastocyst stage embryos, characterized in that: The method comprises: The contour of the trophoblast mask obtained by segmenting the blastocyst stage embryo image is extracted to obtain the trophoblast outer contour and the trophoblast inner contour; Obtaining the center point of the outer contour of the trophoblast, and sampling the outer contour of the trophoblast multiple times to obtain an outer contour sampling point set; Obtaining connecting lines between each outer contour sampling point in the outer contour sampling point set and the center point, and obtaining an inner contour sampling point set based on intersection points between each connecting line and the inner contour of the trophoblast; Based on the distance between the outer contour sampling points and the inner contour sampling points in the same sampling order, the outer and inner point pair distances corresponding to each sampling order are obtained; Map the distances between the outer and inner point pairs corresponding to each sampling order to the coordinate system to obtain a waveform; The total number of peaks in the waveform and the width of each peak are obtained; from the total number of peaks, the number of peaks whose widths do not fall within a preset interval is subtracted; the preset interval is determined based on the number of pixel units occupied by cells; and the subtraction result is used as an estimated number of cells on the trophoblast.

2. The method according to claim 1, characterized in that The outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set, including: The outer contour of the trophoblast is sampled multiple times in a manner of uniformly sampling in 360° azimuth angle to obtain an outer contour sampling point set; or, According to the sampling method of the Douglas-Peucker algorithm, the outer contour of the trophoblast is sampled multiple times to obtain an outer contour sampling point set.

3. The method according to claim 1, characterized in that The method further comprises: The distribution of the cells on the trophoblast is quantified according to the perimeter of the outer contour of the trophoblast and the estimated number of the cells to obtain a quantitative distribution result.

4. The method according to claim 1, characterized in that: The method further comprises: Extracting the contour of the inner cell mass mask in the blastocyst stage embryo image to obtain the inner cell mass contour; Based on the inner cell mass outline, the perimeter and area of ​​the inner cell mass are obtained; The circularity of the inner cell mass was obtained based on the perimeter and area of ​​the inner cell mass.

5. The method according to claim 4, characterized in that The method further comprises: Collect images of the embryo during the blastocyst stage development process, obtain images of the blastocyst stage embryo at multiple imaging moments, and obtain the inner cell mass contour corresponding to each imaging moment; According to the inner cell mass contour corresponding to each imaging moment, the center point position of the inner cell mass at each imaging moment is obtained; The center point position of the inner cell mass at each imaging moment is mapped to the coordinate system to obtain the movement trajectory of the inner cell mass during the blastocyst stage development.

6. The method according to claim 1, characterized in that The method further comprises: Collect images of embryos during the blastocyst stage development process to obtain images of blastocyst stage embryos at multiple imaging moments; Extract the corresponding blastocyst cavity mask and trophoblast outer contour from the blastocyst stage embryo image corresponding to the same imaging moment; According to the image area enclosed by the blastocyst cavity mask and the outer contour of the trophoblast layer corresponding to the same imaging moment, the blastocyst cavity proportion corresponding to each of the multiple imaging moments is obtained; According to the proportion of the blastocyst cavity corresponding to each of the multiple imaging moments, the change trend of the blastocyst cavity is obtained.

7. The method according to claim 1, characterized in that The method further comprises: Acquiring a blastocyst cavity mask extracted from the blastocyst stage embryo image; Obtaining the blastocyst cavity ratio according to the area of ​​the blastocyst cavity mask and the area of ​​the image region enclosed by the outer contour of the trophoblast; Based on the blastocyst cavity ratio, determining whether the embryo is in blastocyst stage III or blastocyst stage IV; Obtaining the center point of the outer contour of the trophoblast layer, and sampling the outer contour of the trophoblast layer multiple times to obtain an outer contour sampling point set, including: if the embryo is in the blastocyst stage III or blastocyst stage IV, obtaining the center point of the outer contour of the trophoblast layer, and sampling the outer contour of the trophoblast layer multiple times to obtain an outer contour sampling point set; or, The contour of the inner cell mass mask in the blastocyst stage embryo image is extracted, including: if the embryo is in the blastocyst stage III or blastocyst stage IV, the contour of the inner cell mass mask in the blastocyst stage embryo image is extracted.

8. A device for analyzing characteristics of blastocyst stage embryos, characterized in that: The device comprises: A contour extraction module is used to extract the contour of the trophoblast mask obtained by segmenting the blastocyst stage embryo image to obtain the trophoblast outer contour and the trophoblast inner contour; A center point acquisition module, used to acquire the center point of the outer contour of the trophoblast; An outer contour sampling module, used for sampling the outer contour of the trophoblast layer multiple times to obtain an outer contour sampling point set; An inner contour sampling module, used to obtain the connecting lines between each outer contour sampling point in the outer contour sampling point set and the center point, and obtain the inner contour sampling point set based on the intersection points between each connecting line and the inner contour of the trophoblast; A distance determination module, for obtaining the distances between the outer contour sampling points and the inner contour sampling points corresponding to each sampling order based on the distances between the outer contour sampling points and the inner contour sampling points in the same sampling order; A mapping analysis module is used to map the distances of the outer and inner point pairs corresponding to each sampling order to a coordinate system to obtain a waveform, and obtain the total number of peaks in the waveform and the width of each peak; from the total number of peaks, subtract the number of peaks whose widths do not fall within a preset interval; the preset interval is determined based on the number of pixel units occupied by cells; and the subtraction result is used as the estimated number of cells on the trophoblast.

9. An embryo analysis system, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images