Cell image analysis method and system based on geometric mean and graph structure

Abstract

The application discloses a cell image analysis method and system based on geometric mean and graph structure, and the method comprises the following steps: inputting a multi-channel fluorescence image into a cell recognition algorithm model to obtain a cell distribution graph; calculating a cell deviation parameter in the cell distribution graph according to a geometric mean algorithm; extracting cell graph features corresponding to the cell distribution graph and the cell deviation parameter based on a graph neural network; and predicting a cell pathology classification and a change trend at a future time point based on an LSTM neural network according to a time sequence corresponding to the cell graph features at a plurality of historical time points. It can be seen that the application can accurately predict the cell pathology classification and the evolution trend at the future time point, improve the automation level and the forward-looking judgment ability of cell pathology analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a cell image analysis method and system based on geometric mean and graph structure. BACKGROUND

[0002] In the prior art, cell pathology analysis mainly relies on manual observation of fluorescence images or uses traditional image processing algorithms to segment and count cell images. However, the traditional method has the problems of poor robustness and limited feature extraction capability when processing multi-channel fluorescence images, and it is difficult to accurately identify the complex distribution relationship between cells. In addition, the existing analysis method generally lacks the modeling and prediction ability of cell distribution trend, and cannot combine historical image data for time series evolution analysis, resulting in a lack of forward-looking judgment of the development of cell pathology state, and it is difficult to meet the actual needs of disease warning and dynamic tracking. It can be seen that the prior art has defects and needs to be solved. SUMMARY

[0003] The technical problem to be solved by the present application is to provide a cell image analysis method and system based on geometric mean and graph structure, which can accurately predict the cell pathology classification and evolution trend at future time points, improve the automation level and forward-looking judgment ability of cell pathology analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0004] To solve the above technical problems, the present application discloses a cell image analysis method based on geometric mean and graph structure in the first aspect, which comprises:

[0005] inputting the multi-channel fluorescence image into a cell recognition algorithm model to obtain a cell distribution map;

[0006] calculating a cell deviation parameter in the cell distribution map according to a geometric mean algorithm;

[0007] extracting cell graph features corresponding to the cell distribution map and the cell deviation parameter based on a graph neural network;

[0008] predicting the cell pathology classification and change trend at future time points based on the time series sequence corresponding to the cell graph features at multiple historical time points based on an LSTM neural network.

[0009] As an optional implementation, in the first aspect of the present application, the inputting the multi-channel fluorescence image into the cell recognition algorithm model to obtain the cell distribution map comprises:

[0010] inputting the cell fluorescence image of each channel into the trained cell image segmentation algorithm model to obtain the cell distribution map in the cell fluorescence image of each channel.

[0011] As an optional implementation, in the first aspect of the present application,

[0012] The cell distribution map comprises a plurality of cell images and a cell type corresponding to each cell image.

[0013] As an optional implementation, in the first aspect of the present application, the calculation of the cell deviation parameter in the cell distribution map according to the geometric mean algorithm comprises:

[0014] For each cell image in the cell distribution map, a geometric mean value of the image deviation parameters of the cell image in all channel cell fluorescence images is calculated to obtain an image average parameter corresponding to the cell image;

[0015] A geometric mean value of the position distance between the image position of the cell image and a plurality of reference cell positions is calculated to obtain a position average parameter corresponding to the cell image;

[0016] A weighted sum average value of the image average parameter and the position average parameter is calculated to obtain a cell deviation parameter corresponding to the cell image.

[0017] As an optional implementation, in the first aspect of the present application, the image deviation parameter is a difference value between an image parameter of the cell image and a standard parameter value; the image parameter comprises at least one of an image brightness, an image color value, an image shape, an image texture and an image spatial distribution parameter.

[0018] As an optional implementation, in the first aspect of the present application, the reference cell position is a position of a cell image of a nearest preset dangerous cell type, a position of a cell image of a same cell type, or a position of a cell image at a center of a preset cell region.

[0019] As an optional implementation, in the first aspect of the present application, the extraction of the cell map features corresponding to the cell distribution map and the cell deviation parameter based on the graph neural network comprises:

[0020] Each cell image in the cell distribution map is taken as a graph node, and the cell deviation parameter corresponding to the cell image corresponding to each graph node is taken as a graph node feature parameter;

[0021] For any two graph nodes, a ratio between the cell deviation parameters corresponding to the two graph nodes respectively is calculated to obtain a node adjacency feature between the two graph nodes;

[0022] All the graph nodes and the corresponding node adjacency features between them are determined as a graph network corresponding to the cell distribution map;

[0023] input the graph network into a trained graph neural network to obtain a cell graph feature corresponding to the graph network; the graph neural network is obtained by training a training data set including a plurality of training cell graph networks and corresponding feature labels.

[0024] As an optional implementation, in the first aspect of the present application, the time sequence of the cell graph features corresponding to a plurality of historical time points is based on an LSTM neural network to predict the cell pathology classification and change trend at a future time point, including:

[0025] The cell graph features of the same patient at a plurality of historical time points are sorted based on time points from early to late to obtain a feature time sequence;

[0026] The feature time sequence is input into a trained LSTM neural network to obtain a plurality of cell pathology types corresponding to the future time point as output;

[0027] According to the type change of the plurality of cell pathology types corresponding to the future time point, the pathological change trend of the patient is determined.

[0028] The second aspect of the embodiment of the present application discloses a cell image analysis system based on geometric mean and graph structure, the system comprises:

[0029] The recognition module is configured to input the multi-channel fluorescence image into a cell recognition algorithm model to obtain a cell distribution map;

[0030] The calculation module is configured to calculate a cell deviation parameter in the cell distribution map according to a geometric mean algorithm;

[0031] The extraction module is configured to extract a cell graph feature corresponding to the cell distribution map and the cell deviation parameter based on a graph neural network;

[0032] The prediction module is configured to predict a cell pathology classification and change trend at a future time point based on an LSTM neural network according to a time sequence of cell graph features corresponding to a plurality of historical time points.

[0033] As an optional implementation, in the second aspect of the present application, the specific manner in which the recognition module inputs the multi-channel fluorescence image into the cell recognition algorithm model to obtain the cell distribution map includes:

[0034] The cell fluorescence image of each channel is input into a trained cell image segmentation algorithm model to obtain a cell distribution map in the cell fluorescence image of each channel.

[0035] As an optional implementation, in the second aspect of the present application,

[0036] The cell distribution map comprises a plurality of cell images and a cell type corresponding to each cell image.

[0037] As an optional implementation, in the second aspect of the present application, the specific manner in which the calculation module calculates the cell deviation parameter in the cell distribution map according to the geometric mean algorithm comprises:

[0038] For each cell image in the cell distribution map, a geometric mean value of image deviation parameters of the cell image in all channel cell fluorescence images is calculated to obtain an image average parameter corresponding to the cell image;

[0039] A geometric mean value of position distances between the image position of the cell image and a plurality of reference cell positions is calculated to obtain a position average parameter corresponding to the cell image;

[0040] A weighted sum average value of the image average parameter and the position average parameter is calculated to obtain a cell deviation parameter corresponding to the cell image.

[0041] As an optional implementation, in the second aspect of the present application, the image deviation parameter is a difference value between an image parameter of the cell image and a standard parameter value; the image parameter comprises at least one of an image brightness, an image color value, an image shape, an image texture and an image spatial distribution parameter.

[0042] As an optional implementation, in the second aspect of the present application, the reference cell position is a position of a cell image of a nearest preset dangerous cell type, a position of a cell image of a nearest same cell type or a position of a cell image at a center of a preset cell region.

[0043] As an optional implementation, in the second aspect of the present application, the specific manner in which the extraction module extracts the cell graph feature corresponding to the cell distribution map and the cell deviation parameter based on a graph neural network comprises:

[0044] Each cell image in the cell distribution map is taken as a graph node, and the cell deviation parameter corresponding to the cell image corresponding to each graph node is taken as a graph node feature parameter;

[0045] For any two graph nodes, a ratio between the cell deviation parameters corresponding to the two graph nodes respectively is calculated to obtain a node adjacency feature between the two graph nodes;

[0046] All the graph nodes and the node adjacency features corresponding to each other are determined as a graph network corresponding to the cell distribution map;

[0047] inputting the graph network into a trained graph neural network to obtain a cell graph feature corresponding to the graph network; the graph neural network is obtained by training a training data set including a plurality of training cell graph networks and corresponding feature labels.

[0048] As an optional implementation, in the second aspect of the present application, the prediction module predicts the cell pathology classification and the specific manner of the change trend of the future time point based on an LSTM neural network according to a time sequence sequence corresponding to the cell graph features of a plurality of historical time points, and the specific manner includes:

[0049] sequencing the cell graph features of the same patient at a plurality of historical time points based on time points from early to late to obtain a feature time sequence;

[0050] inputting the feature time sequence into a trained LSTM neural network to obtain a plurality of cell pathology types corresponding to the future time point as output;

[0051] determining the pathological change trend of the patient according to the type change of the plurality of cell pathology types corresponding to the future time point.

[0052] The third aspect of the present application discloses another cell image analysis system based on geometric mean and graph structure, and the system includes:

[0053] a memory storing executable program codes;

[0054] a processor coupled with the memory;

[0055] The processor invokes the executable program codes stored in the memory to execute part or all of the steps of the cell image analysis method based on geometric mean and graph structure disclosed in the first aspect of the present application.

[0056] The fourth aspect of the present application discloses a computer storage medium storing computer instructions, and the computer instructions are used to execute part or all of the steps of the cell image analysis method based on geometric mean and graph structure disclosed in the first aspect of the present application when invoked.

[0057] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0058] The present application obtains a cell distribution map by inputting a multi-channel fluorescence image into a cell recognition algorithm model, and calculates a cell deviation degree parameter in the map by using a geometric mean algorithm, and on this basis, combines a graph neural network to extract a graph structure feature of the fused cell distribution map and deviation degree parameter, and inputs the same into an LSTM neural network through construction of a time sequence to perform cell pathology classification and change trend prediction, so that accurate prediction of cell pathology classification and evolution trend at a future time point can be realized, the automation level and forward-looking judgment capability of cell pathology analysis are improved, and efficient and reliable technical support is provided for disease monitoring and auxiliary diagnosis. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0060] Figure 1 is a flow diagram of a cell image analysis method based on geometric mean and graph structure disclosed by the embodiments of the present application.

[0061] Figure 2 is a structural diagram of a cell image analysis system based on geometric mean and graph structure disclosed by the embodiments of the present application.

[0062] Figure 3 is another structural diagram of a cell image analysis system based on geometric mean and graph structure disclosed by the embodiments of the present application. DETAILED DESCRIPTION

[0063] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0064] The terms "first", "second", etc. in the specification and claims of the present application and the above drawings are used to distinguish different objects, not to describe a specific order. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or equipment including a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed, or can optionally include other steps or units inherent to the process, method, product or equipment.

[0065] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily mutually exclusive of one another. As will be apparent to those of ordinary skill in the art, embodiments described herein can be combinable with other embodiments.

[0066] The application discloses a cell image analysis method and system based on geometric mean and graph structure, obtains a cell distribution graph by inputting a multi-channel fluorescence image into a cell recognition algorithm model, calculates a cell deviation degree parameter in the graph by using a geometric mean algorithm, extracts a graph structure feature of the cell distribution graph and the deviation degree parameter in combination with a graph neural network on the basis, and inputs the graph structure feature into an LSTM neural network through construction of a time sequence to perform cell pathology classification and change trend prediction, so that accurate prediction of cell pathology classification and evolution trend at a future time point can be realized, the automation level and the forward-looking judgment capability of cell pathology analysis are improved, and efficient and reliable technical support is provided for disease monitoring and auxiliary diagnosis. The following will be described in detail.

[0067] Embodiment one

[0068] Please refer to Figure 1 , Figure 1 is a flowchart of a cell image analysis method based on geometric mean and graph structure disclosed by the embodiment of the application. Wherein, Figure 1 The cell image analysis method based on geometric mean and graph structure described above can be applied in a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). As Figure 1 shown, the cell image analysis method based on geometric mean and graph structure can include the following operations:

[0069] 101, input a multi-channel fluorescence image into a cell recognition algorithm model to obtain a cell distribution graph.

[0070] 102, calculate a cell deviation degree parameter in the cell distribution graph according to a geometric mean algorithm.

[0071] 103, extract a cell graph feature corresponding to the cell distribution graph and the cell deviation degree parameter based on a graph neural network.

[0072] 104, predict cell pathology classification and change trend at a future time point based on an LSTM neural network according to a time sequence corresponding to a plurality of historical time point cell graph features.

[0073] It can be seen that the above embodiment of the application can realize accurate prediction of cell pathological classification and evolution trend at a future time point, improve the automation level and forward-looking judgment ability of cell pathological analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0074] As an optional embodiment, in the above step of inputting the multi-channel fluorescence image into the cell recognition algorithm model to obtain the cell distribution map, the step includes:

[0075] The cell fluorescence image of each channel is input into the trained cell image segmentation algorithm model to obtain the cell distribution map in the cell fluorescence image of each channel.

[0076] It can be seen that, by the above optional embodiment, the cell fluorescence image of each channel is input into the trained cell image segmentation algorithm model, the cell region and distribution information corresponding to each channel can be accurately extracted from the original fluorescence image, not only the accuracy and robustness of cell image segmentation are improved, but also the recognition error caused by image overlap and signal interference in the traditional method in multi-channel image processing is avoided, thereby providing a clear structure and channel-independent image basis for subsequent cell morphology analysis and state recognition, and effectively improving the overall image processing efficiency and analysis accuracy.

[0077] As an optional embodiment, in the above step of inputting the multi-channel fluorescence image into the cell recognition algorithm model to obtain the cell distribution map, the step includes:

[0078] The cell distribution map includes a plurality of cell images and a cell type corresponding to each cell image.

[0079] It can be seen that, by the above optional embodiment, the content of the cell distribution map is limited to comprehensively represent the cell distribution characteristics, to assist in realizing accurate prediction of cell pathological classification and evolution trend at a future time point, improve the automation level and forward-looking judgment ability of cell pathological analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0080] As an optional embodiment, in the above step of calculating the cell deviation parameter in the cell distribution map according to the geometric mean algorithm, the step includes:

[0081] For each cell image in the cell distribution map, the geometric mean value of the image deviation parameters of the cell image in the cell fluorescence images of all channels is calculated to obtain an image average parameter corresponding to the cell image.

[0082] calculating a geometric mean of position distances between the image position of the cell image and a plurality of reference cell positions, to obtain a position average parameter corresponding to the cell image;

[0083] calculating a weighted sum average of the image average parameter and the position average parameter, to obtain a cell deviation parameter corresponding to the cell image.

[0084] It can be seen that, through the above optional embodiments, for each cell image in the cell distribution map, the geometric mean of the image deviation parameters of the cell image in the cell fluorescence images of all channels is calculated, which can uniformly measure the overall deviation of the same cell in the multi-channel image in terms of morphology or signal expression; further, the geometric mean of the position distances between the position of the cell image and a plurality of reference cell positions is calculated to evaluate the outlying degree of the spatial distribution; finally, the image average parameter and the position average parameter are weighted and summed to obtain a more comprehensive representative cell deviation parameter, which realizes the fusion quantification of the deviation degree of a single cell in the image signal and the spatial position two dimensions, and provides key index support for abnormal cell recognition and subsequent pathological trend modeling.

[0085] As an optional embodiment, in the above steps, the image deviation parameter is a difference value between the image parameter of the cell image and a standard parameter value; the image parameter includes at least one of image brightness, image color value, image morphology, image texture, and image spatial distribution parameter.

[0086] It can be seen that, through the above optional embodiments, the details of the image deviation parameter are limited to facilitate subsequent accurate characterization of the image deviation parameter of the cell, to assist in accurately predicting the pathological classification of the cell at a future time point and its evolution trend, to improve the automation level and forward-looking judgment ability of cell pathological analysis, and to provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0087] As an optional embodiment, in the above steps, the reference cell position is a position of a cell image of a nearest preset dangerous cell type, a position of a cell image of a nearest same cell type, or a position of a cell image at a center of a preset cell region.

[0088] It can be seen that, through the above optional embodiments, the type of the reference cell position is limited to facilitate subsequent accurate calculation of the position average parameter of the cell, to assist in accurately predicting the pathological classification of the cell at a future time point and its evolution trend, to improve the automation level and forward-looking judgment ability of cell pathological analysis, and to provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0089] As an optional embodiment, in the above steps, based on a graph neural network, cell image features corresponding to the cell distribution map and the cell deviation parameter are extracted, including:

[0090] each cell image in the cell distribution map as a graph node, and a cell deviation parameter corresponding to each graph node as a graph node feature parameter;

[0091] For any two graph nodes, a ratio between the cell deviation parameters corresponding to the two graph nodes is calculated to obtain a node adjacency feature between the two graph nodes.

[0092] All graph nodes and the corresponding node adjacency features therebetween are determined as a graph network corresponding to the cell distribution map.

[0093] The graph network is input into a trained graph neural network to obtain cell map features corresponding to the graph network. The graph neural network is trained by a training data set including a plurality of training cell map networks and corresponding feature labels.

[0094] As can be seen, by the above optional embodiments, each cell image in the cell distribution map is taken as a graph node, and a cell deviation parameter corresponding to each graph node is taken as a graph node feature parameter, which can fuse the morphological and positional information of single cells in the cell map structure. Further, by calculating the ratio between the cell deviation parameters of any two graph nodes, the node adjacency features therebetween are extracted, which helps to depict the relative differences between cells in terms of deviation degree. Furthermore, by constructing a cell graph network and inputting it into a trained graph neural network, deep features of cell population distribution patterns can be effectively extracted, the comprehensive representation of cell state in the spatial structure dimension is realized, and the sensitivity and recognition ability for local abnormalities and pathological trends are improved.

[0095] As an optional embodiment, in the above step, predicting the cell pathology classification and change trend at the future time point based on the time sequence of the cell map features at the plurality of historical time points includes:

[0096] The cell map features of the same patient at the plurality of historical time points are sorted based on the time points from early to late to obtain a feature time sequence.

[0097] The feature time sequence is input into a trained LSTM neural network to obtain a plurality of cell pathology types corresponding to the future time point as output.

[0098] According to the type change of the plurality of cell pathology types corresponding to the future time point, a pathological change trend corresponding to the patient is determined.

[0099] As can be seen, by the above optional embodiments, through continuous modeling of time sequence features and trend prediction based on deep neural networks, the accurate recognition ability of pathological evolution trend and the dynamic adaptability of the model to the pathological development process are improved, and the forward-looking and reliability of cell pathology change analysis are enhanced.

[0100] Embodiment Two

[0101] Referring to Figure 2 , Figure 2 is a structural schematic diagram of a cell image analysis system based on geometric mean and graph structure according to an embodiment of the present application. Wherein, Figure 2 The cell image analysis system based on geometric mean and graph structure described can be applied in a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). As Figure 2 shown, the cell image analysis system based on geometric mean and graph structure can include:

[0102] The recognition module 201 is configured to input the multi-channel fluorescence image into a cell recognition algorithm model to obtain a cell distribution map.

[0103] The calculation module 202 is configured to calculate a cell deviation parameter in the cell distribution map according to a geometric mean algorithm.

[0104] The extraction module 203 is configured to extract a cell graph feature corresponding to the cell distribution map and the cell deviation parameter based on a graph neural network.

[0105] The prediction module 204 is configured to predict a cell pathology classification and a change trend at a future time point based on an LSTM neural network according to a time sequence sequence corresponding to cell graph features at a plurality of historical time points.

[0106] As can be seen, the above-mentioned embodiment of the application obtains a cell distribution map by inputting a multi-channel fluorescence image into a cell recognition algorithm model, and calculates a cell deviation parameter in the map using a geometric mean algorithm. On this basis, a graph structure feature fusing the cell distribution map and the deviation parameter is extracted by combining a graph neural network, and a cell pathology classification and a change trend are predicted by inputting a time sequence sequence into an LSTM neural network, so as to realize accurate prediction of a cell pathology classification and an evolution trend at a future time point, improve the automation level and the forward-looking judgment ability of cell pathology analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0107] As an optional embodiment, the specific manner in which the recognition module inputs the multi-channel fluorescence image into the cell recognition algorithm model to obtain the cell distribution map includes:

[0108] The cell fluorescence image of each channel is input into a trained cell image segmentation algorithm model to obtain a cell distribution map in the cell fluorescence image of each channel.

[0109] It can be seen that by the above optional embodiment, the cell fluorescence image of each channel is input into the trained cell image segmentation algorithm model, the cell region and distribution information corresponding to each channel can be accurately extracted from the original fluorescence image, the accuracy and robustness of cell image segmentation are improved, and the recognition error caused by image overlap and signal interference in the traditional method in multi-channel image processing is avoided, thereby providing a clear structure and channel-independent image basis for subsequent cell morphology analysis and state recognition, and effectively improving the overall image processing efficiency and analysis accuracy.

[0110] As an optional embodiment,

[0111] The cell distribution map includes a plurality of cell images and a cell type corresponding to each cell image.

[0112] It can be seen that by the above optional embodiment, the content of the cell distribution map is limited to fully characterize the cell distribution characteristics, to assist in accurately predicting the cell pathology classification and evolution trend at the future time point, to improve the automation level and forward-looking judgment ability of cell pathology analysis, and to provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0113] As an optional embodiment, the specific manner in which the calculation module calculates the cell deviation parameter in the cell distribution map according to the geometric mean algorithm includes:

[0114] For each cell image in the cell distribution map, the geometric mean of the image deviation parameters of the cell image in the cell fluorescence images of all channels is calculated to obtain an image average parameter corresponding to the cell image;

[0115] The geometric mean of the position distance between the image position of the cell image and a plurality of reference cell positions is calculated to obtain a position average parameter corresponding to the cell image;

[0116] The weighted sum average of the image average parameter and the position average parameter is calculated to obtain a cell deviation parameter corresponding to the cell image.

[0117] It can be seen that, through the above optional embodiments, for each cell image in the cell distribution map, the geometric mean of the image deviation parameters of the cell image in all channel cell fluorescence images is calculated, which can uniformly measure the overall deviation of the same cell in the multi-channel image in terms of morphology or signal expression; further, by calculating the geometric mean of the position distance between the position of the cell image and a plurality of reference cell positions, the outlying degree of the spatial distribution thereof is evaluated; finally, the image average parameter and the position average parameter are weighted and summed to obtain a more comprehensive representative cell deviation degree parameter, which realizes the fusion quantification of the deviation degree of a single cell in the image signal and the spatial position two dimensions, and provides key index support for abnormal cell identification and subsequent pathological trend modeling.

[0118] As an optional embodiment, the image deviation parameter is the difference between the image parameter of the cell image and the standard parameter value; the image parameter includes at least one of image brightness, image color value, image morphology, image texture, and image spatial distribution parameter.

[0119] It can be seen that, through the above optional embodiments, the details of the image deviation parameter are defined to facilitate subsequent accurate characterization of the image deviation parameter of the cell, assist in realizing accurate prediction of the pathological classification of the cell at a future time point and its evolution trend, improve the automation level and forward-looking judgment ability of cell pathological analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0120] As an optional embodiment, the reference cell position is the position of the cell image of the nearest preset dangerous cell type, the position of the cell image of the nearest same cell type, or the position of the cell image at the center of the preset cell region.

[0121] It can be seen that, through the above optional embodiments, the type of the reference cell position is defined to facilitate subsequent accurate calculation of the position average parameter of the cell, assist in realizing accurate prediction of the pathological classification of the cell at a future time point and its evolution trend, improve the automation level and forward-looking judgment ability of cell pathological analysis, and provide efficient and reliable technical support for disease monitoring and auxiliary diagnosis.

[0122] As an optional embodiment, the extraction module extracts the specific manner of the cell graph feature corresponding to the cell distribution map and the cell deviation degree parameter based on a graph neural network, which includes:

[0123] Each cell image in the cell distribution map is taken as a graph node, and the cell deviation degree parameter corresponding to the cell image corresponding to each graph node is taken as a graph node feature parameter;

[0124] For any two graph nodes, the ratio between the cell deviation degree parameters corresponding to the two graph nodes is calculated to obtain the node adjacency feature between the two graph nodes.

[0125] All the graph nodes and the corresponding node adjacency features between them are determined as the graph network corresponding to the cell distribution graph;

[0126] The graph network is input into the trained graph neural network to obtain the cell graph features corresponding to the graph network; the graph neural network is obtained by training the training data set including a plurality of training cell graph networks and corresponding feature labels.

[0127] As can be seen, through the above optional embodiments, each cell image in the cell distribution graph is taken as a graph node, and the cell deviation degree parameter corresponding to each graph node is taken as a graph node feature parameter, which can fuse the morphological and positional information of single cells in the cell graph structure; further, by calculating the ratio between the cell deviation degree parameters of any two graph nodes, the node adjacency features between them are extracted, which helps to depict the relative difference between cells in terms of deviation degree; then, the cell graph network is constructed and input into the trained graph neural network, which can effectively extract the deep features of the cell population distribution pattern, realize the comprehensive representation of the cell state in the spatial structure dimension, and improve the sensitivity and recognition ability of local abnormalities and pathological trends.

[0128] As an optional embodiment, the prediction module predicts the cell pathology classification and change trend of the future time point based on the LSTM neural network according to the time sequence sequence of the cell graph features of a plurality of historical time points, and the specific manner includes:

[0129] The cell graph features of the same patient at a plurality of historical time points are sorted based on the time points from early to late to obtain a feature time sequence;

[0130] The feature time sequence is input into the trained LSTM neural network to obtain the output of the plurality of cell pathology types corresponding to the future time point;

[0131] According to the type change of the plurality of cell pathology types corresponding to the future time point, the pathological change trend of the patient is determined.

[0132] As can be seen, through the above optional embodiments, through the continuous modeling of time sequence features and the trend prediction based on deep neural networks, the accurate recognition ability of the pathological evolution trend and the dynamic adaptability of the model to the pathological development process are improved, and the forward-looking and reliability of the cell pathology change analysis are enhanced.

[0133] Embodiment three

[0134] Please refer to Figure 3 , Figure 3 is another cell image analysis system based on geometric mean and graph structure disclosed in the embodiments of the present application. Figure 3The described geometric mean and graph structure based cell image analysis system is applied in a data processing system / data processing device / data processing server (wherein the server comprises a local processing server or a cloud processing server). As shown in Figure 3 The geometric mean and graph structure based cell image analysis system can comprise:

[0135] a memory 301 storing executable program code;

[0136] a processor 302 coupled with the memory 301;

[0137] The processor 302 invokes the executable program code stored in the memory 301 to perform the steps of the geometric mean and graph structure based cell image analysis method described in Embodiment One.

[0138] Embodiment Four

[0139] The embodiments of the present application disclose a computer readable storage medium storing a computer program for electronic data exchange, wherein the computer program causes a computer to perform the steps of the geometric mean and graph structure based cell image analysis method described in Embodiment One.

[0140] Embodiment Five

[0141] The embodiments of the present application disclose a computer program product comprising a non-transitory computer readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform the steps of the geometric mean and graph structure based cell image analysis method described in Embodiment One.

[0142] The above describes certain embodiments of the present application, and other embodiments are within the scope of the appended claims. In some cases, actions recited in the claims can be performed in a different order than the order in which the actions are recited in the embodiments, and still achieve the desired results. In addition, the processes depicted in the figures do not necessarily have to be performed in the specific order depicted in the figures, nor in sequential order, nor all described steps can be performed, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

[0143] The systems, apparatuses, modules, or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0144] For the sake of description, the above-described apparatus is described with various units in function to describe the embodiment. Of course, the functions of the units can be implemented in one or more software and / or hardware in implementing the specification.

[0145] Those skilled in the art should understand that the embodiments of the specification can be provided as a method, a system, or a computer program product. Therefore, the embodiments of the specification can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the specification can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0146] The specification is described with reference to flowcharts and / or block diagrams of methods, apparatus (system), and computer program products according to embodiments of the specification. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing apparatus generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0147] These computer program instructions can also be stored in a computer-readable memory that can direct the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0148] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer-implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that implements the functions specified in the flowcharts and / or block diagrams.

[0149] In a typical configuration, the computing device includes one or more processors (CPU), input / output interface, network interface, and memory.

[0150] Memory can include, without being limited to, non- transitory computer-readable media, random access memory (RAM), and / or read-only memory (ROM), such as flash memory, among others. Memory is an example of computer-readable media.

[0151] Computer-readable media includes permanent and non- permanent, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules, or other data. Examples of computer-readable media include, without being limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disks (DVDs) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. According to the definition provided herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0152] It is also important to note that the terms "comprises", "comprising", or other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0153] The description can be described in the general context of computer- executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, and the like, that perform particular tasks or implement particular abstract data types. Computer-executable instructions, associated data structures, and program modules represent examples of the program code means for executing steps of the methods disclosed herein. The manner in which the computer-executable instructions are executed includes the use of a

[0154] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0155] Finally, it should be noted that the cell image analysis method and system based on geometric mean and graph structure disclosed in the embodiments of the present invention are only preferred embodiments of the present invention, and are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A cell image analysis method based on geometric mean and graph structure, characterized in that: The method comprises: Inputting the multi-channel fluorescence image into a cell recognition algorithm model to obtain a cell distribution map; the cell distribution map includes multiple cell images and the cell type corresponding to each cell image; Calculating the cell deviation parameter in the cell distribution map according to the geometric mean algorithm includes: For each cell image in the cell distribution map, calculating the geometric mean of image deviation parameters of the cell image in the cell fluorescence images of all channels to obtain an image average parameter corresponding to the cell image; the image deviation parameter is the difference between the image parameter of the cell image and the standard parameter value; the image parameter includes at least one of image brightness, image color value, image morphology, image texture and image spatial distribution parameter; Calculating the geometric mean of the positional distances between the image position of the cell image and a plurality of reference cell positions to obtain a positional average parameter corresponding to the cell image; the reference cell position is the position of the nearest cell image of a preset dangerous cell type, the position of the nearest cell image of the same cell type, or the position of a cell image at the center of a preset cell region; Calculating a weighted average of the image average parameter and the position average parameter to obtain a cell deviation parameter corresponding to the cell image; Extracting cell map features corresponding to the cell distribution map and the cell deviation parameter based on a graph neural network; According to the time series corresponding to the cell map features at multiple historical time points, based on the LSTM neural network, the cell pathology classification and change trend at future time points are predicted.

2. The cell image analysis method based on geometric mean and graph structure according to claim 1, characterized in that: The multi-channel fluorescence image is input into the cell recognition algorithm model to obtain a cell distribution map, including: The cell fluorescence image of each channel is input into the trained cell image segmentation algorithm model to obtain the cell distribution map in the cell fluorescence image of each channel.

3. The cell image analysis method based on geometric mean and graph structure according to claim 1, characterized in that: The extracting of cell map features corresponding to the cell distribution map and the cell deviation parameter based on the graph neural network includes: Each cell image in the cell distribution map is a map node, and the cell deviation parameter corresponding to the cell image corresponding to each map node is a map node characteristic parameter; For any two of the graph nodes, calculating the ratio between the cell deviation parameters corresponding to the two graph nodes, and obtaining the node adjacency feature between the two graph nodes; Determine all the graph nodes and the corresponding node adjacency features therebetween as a graph network corresponding to the cell distribution graph; The graph network is input into a trained graph neural network to obtain cell graph features corresponding to the graph network; the graph neural network is trained by a training data set including multiple training cell graph networks and corresponding feature annotations.

4. The cell image analysis method based on geometric mean and graph structure according to claim 1, characterized in that: The method of predicting the cell pathology classification and change trend at a future time point based on the time series corresponding to the cell map features at multiple historical time points and the LSTM neural network includes: Sort the cell map features of the same patient at multiple historical time points from early to late based on the time points to obtain a feature time series sequence; Inputting the characteristic time series sequence into the trained LSTM neural network to obtain multiple cell pathology types corresponding to the output future time points; The pathological change trend corresponding to the patient is determined according to the type changes of the multiple cell pathological types corresponding to the future time point.

5. A cell image analysis system based on geometric mean and graph structure, characterized in that: The system comprises: An identification module is used to input the multi-channel fluorescence image into a cell identification algorithm model to obtain a cell distribution map; the cell distribution map includes multiple cell images and the cell type corresponding to each cell image; A calculation module, configured to calculate the cell deviation parameter in the cell distribution map according to a geometric mean algorithm, comprising: For each cell image in the cell distribution map, calculating the geometric mean of image deviation parameters of the cell image in the cell fluorescence images of all channels to obtain an image average parameter corresponding to the cell image; the image deviation parameter is the difference between the image parameter of the cell image and the standard parameter value; the image parameter includes at least one of image brightness, image color value, image morphology, image texture and image spatial distribution parameter; Calculating the geometric mean of the positional distances between the image position of the cell image and a plurality of reference cell positions to obtain a positional average parameter corresponding to the cell image; the reference cell position is the position of the nearest cell image of a preset dangerous cell type, the position of the nearest cell image of the same cell type, or the position of a cell image at the center of a preset cell region; Calculating a weighted average of the image average parameter and the position average parameter to obtain a cell deviation parameter corresponding to the cell image; An extraction module, configured to extract cell map features corresponding to the cell distribution map and the cell deviation parameter based on a graph neural network; The prediction module is used to predict the cell pathology classification and change trend at future time points based on the time series corresponding to the cell map features at multiple historical time points and based on the LSTM neural network.

6. A cell image analysis system based on geometric mean and graph structure, characterized in that: The system comprises: a memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the cell image analysis method based on geometric mean and graph structure according to any one of claims 1 to 4.

Images