Tissue imaging processing method, system, and computer-readable recording medium

Abstract

To relate to the field of image processing technology and provide a method and system for processing tissue images on whole-slide imaging.SOLUTION: A method includes: obtaining a target image; obtaining a dataset; enhancing the staining components of the dataset; using the enhanced dataset for federated learning to construct a tumor segmentation model, the dataset including several test images, and the test images being immunohistochemical images; inputting the target image into the tumor segmentation model to determine a tumor area within the target image; performing cell membrane staining detection on the tumor area; classifying the tumor area based on the integrity of the cell membrane and the level of staining; and grading based on the classification result.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to the field of image processing techniques, and more particularly to methods and systems for processing tissue images in whole slide imaging. [Background technology]

[0002] The rapid development of precision medicine and personalized medicine has increased the demand for accurate diagnostic tests. These tests are crucial for implementing new treatments, such as targeted therapies. With existing technologies, target tissue image processing relies heavily on manual identification, which relies on human experience, to determine gene expression levels in target tissue images. This method generally lacks precision, making it an urgent task to improve the accuracy of image processing. Summary of the Invention [Problem to be solved by the invention]

[0003] In this regard, the present disclosure provides a method and system for processing tissue images in whole slide imaging, which aims to optimize the method and evaluate the level or expression of human epidermal growth factor receptor 2 (HER2) in target tissue images more accurately than conventional techniques. [Means for solving the problem]

[0004] In a first aspect, the present disclosure provides a specific embodiment of a method for processing tissue images from whole slide imaging, the method including: (a) acquiring a target image; (b) acquiring a dataset, enhancing staining components of the dataset, and using the enhanced dataset for associative learning to construct a tumor segmentation model, the dataset including a test image, the test image being an immunohistochemical (IHC) image; (c) inputting the target image into the tumor segmentation model to determine a tumor region of the target image; (d) performing cell membrane staining detection on the tumor region, and further classifying the tumor region based on the integrity of the cell membrane and the level of staining of the cell membrane, and ranking the tumor region based on the classification result.

[0005] In some embodiments, "acquiring a dataset and enhancing staining components of the dataset" further includes acquiring a test image of the dataset, performing color separation on the test image, obtaining a color base and staining intensity of a dye in the test image, randomly scaling and randomly translating the staining intensity of the dye to enhance the staining intensity of the dye, randomly rotating the color base of the dye to obtain an enhanced color base of the dye, and integrating the enhanced color base and the enhanced staining intensity of the dye to complete the enhancement of the staining component.

[0006] In some embodiments, "using the enhanced datasets for federated learning to construct the tumor segmentation model" further comprises preprocessing some of the enhanced datasets and inputting preprocessing results into a federated learning server to construct the tumor segmentation model. Preprocessing is used to reduce the amount of data input into a federated learning server.

[0007] In some embodiments, "preprocessing the highlighted datasets and inputting the preprocessing results into a federated learning server to construct the tumor segmentation model" further includes constructing a local model and a global model based on the learning rules of the federated learning server, training the local model on the highlighted one or more datasets to obtain parameters of the local model, inputting the parameters of the local model into the federated learning server, training the global model through a clustering pipeline, training the global model on the highlighted datasets, inputting the parameters of the trained global model into the federated learning server, and constructing the tumor segmentation model.

[0008] In some embodiments, (d) "performing cell membrane staining detection on the tumor region and classifying the tumor region" further includes: (d1) detecting nuclei in the tumor region to determine the positions of the nuclei in the tumor region; (d2) performing cell membrane detection based on the positions of the nuclei and determining the positions of the cell membranes corresponding to the nuclei; (d3) classifying the cell membranes based on the staining level and obtaining a pre-classification result; and (d4) classifying the tumor region based on the pre-classification result and the integrity of the cell membrane.

[0009] In some embodiments, (d2) "performing cell membrane detection based on the position of the nucleus and determining the position of the cell membrane corresponding to the nucleus" further includes expanding a segmentation mask of the nucleus based on the position of the nucleus and determining the position of the cell membrane corresponding to the nucleus.

[0010] (d3) "Classifying the cell membrane based on the staining level and obtaining a pre-classification result" further includes performing color separation on each pixel of the cell membrane and converting it into a representation in hematoxylin (H&E)-diaminobenzidine (DAB), classifying each pixel of the cell membrane according to a preset threshold based on the staining intensity of the diaminobenzidine channel, obtaining the classification level of each pixel, and obtaining the pre-classification result based on the classification level of each pixel.

[0011] In some embodiments, (d4) "classifying the tumor region based on the pre-classification result and the cell membrane integrity" further comprises using a skeletonization algorithm to obtain the cell membrane integrity and refining the pre-classification result based on the cell membrane integrity to classify the tumor region.

[0012] In some embodiments, grading based on the classification results includes assessing the expression level of HER2 in the target image using predetermined standards, such as the ASCO CAP 2018 guidelines.

[0013] In a second aspect, the present disclosure also provides a specific embodiment of a tissue image processing system for whole slide imaging, the system including a memory, a processor, and a computer program stored in the memory, the processor executing the computer program to implement the tissue image processing method of the first aspect. .

[0014] In a third aspect, the present disclosure also provides certain embodiments of a computer program product, the computer program product including computer programs / instructions that, when executed by a processor, implement the tissue imaging processing method of the first embodiment. [Effects of the Invention]

[0015] Compared with the prior art, the present invention has at least the following beneficial effects or unexpected results: 1. The expression of HER2 in the target image can be accurately evaluated. 2. Target images obtained from different facilities can be optimized, enabling cross-facility model learning without compromising user privacy. [Brief explanation of the drawings]

[0016] [Figure 1] FIG. 1 is a schematic diagram of a whole-slide tissue imaging method according to an embodiment of the present invention. [Figure 2] FIG. 2 is another schematic diagram of a whole-slide tissue imaging method according to an embodiment of the present invention. [Figure 3] FIG. 3 is another schematic diagram of a whole-slide tissue imaging method according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] One or more specific embodiments are illustrated in the accompanying drawings by way of example, and not by way of limitation. The drawings are not to scale unless specifically disclosed. The present disclosure should be understood by those of ordinary skill in the art in light of the following detailed description of preferred specific embodiments and with reference to the accompanying drawings.

[0018] The above-mentioned characteristics, technical features, advantages, and realizations thereof of the present disclosure will be further explained below in a clear and understandable manner in conjunction with examples of preferred specific embodiments.

[0019] In order to more clearly explain the technical solutions of the embodiments of the present application, specific embodiments will be described below with reference to the accompanying drawings. The drawings described below are only a part of the embodiments of the present application. Those skilled in the art can obtain other drawings and other embodiments based on these drawings without any creative effort. Any modifications and improvements made without departing from the spirit of the present application are within the protection scope of the present application.

[0020] In order to simplify the drawings, only parts related to the corresponding embodiments are illustrated in a schematic manner. These do not represent the actual structure of the product. Furthermore, in order to simplify the drawings and make them easier to understand, some components having the same structure or function are only partially illustrated in a schematic manner. In reality, there may be more or fewer components having the same structure or function.

[0021] In this disclosure, unless expressly stated otherwise, ordinal numbers such as "first" and "second" are used to distinguish related objects and do not indicate relative importance or order. "Plurality" includes two or more, as well as other quantifiers. " / " is used to express a relationship between related objects and indicates an "or" relationship. "And / or" includes any combination relationship between related objects, such as "a and / or b," and is used to describe a relationship between related objects. "a and / or b" includes "a alone," "b alone," or "a and b." "One or more" or "at least one" of multiple objects means any combination of any one or more objects, such as "one or more of a1, a2, a3" or "at least one of a1, a2, a3." "One or more of a1, a2, and a3" or "at least one of a1, a2, and a3" includes "a1 alone," "a2 alone," "a3 alone," "a1 and a2," "a1 and a3," "a2 and a3," or "a1, a2, and a3."

[0022] Recent advances in precision oncology, such as immunotherapy and antibody-drug conjugates, have shown great potential for extending cancer patient survival. To develop appropriate treatment strategies, these advances are often accompanied by meticulous investigation of patient information, such as high-throughput sequencing and companion diagnostics. Such tests greatly benefit from artificial intelligence (AI) and deep learning (DL) models that provide quantitative details of biomarker expression.

[0023] Human epidermal growth factor receptor 2 (HER2) plays an important role in protein-targeted therapy in the treatment of metastatic breast cancer. Trastuzumab (Herceptin), a HER2-targeted monoclonal antibody drug, has matured as a first-line treatment for HER2-positive patients, significantly reducing the risk of disease progression and death compared with chemotherapy alone. HER2 expression is assessed by immunohistochemical (IHC) staining and, if the former is considered equivocal, in situ hybridization testing. However, recent studies have shown that clinical HER2 interpretation may be compromised by low interpathologist consistency and that interpathologist agreement can be improved with the aid of several artificial intelligence algorithms. Clearly, the application of AI in HER2 expression analysis is urgently needed.

[0024] The development of AI systems for HER2 expression assessment still faces several challenges. First, deep learning models integrated into such systems must be data-driven, yet medical images are often a scarce resource for a single institution. More critically, patient privacy policies prohibit data sharing between institutions, making it impossible to train deep learning models using traditional methods. Fortunately, the emerging federated learning (FL) strategy allows medical institutions to collaboratively develop deep learning models while maintaining patient privacy. Specifically, in the FL strategy, multiple institutions train models in parallel using their respective data and pass model parameters between each other during iterations. This approach has proven to significantly improve the generalizability of various deep learning models for medical imaging.

[0025] Second, while FL applications have been successful in computed tomography and magnetic resonance imaging, the performance of DL models specifically for histological images can be hindered by imaging variations across laboratories. Histological images, such as IHC stains required for HER2 expression assessment, consist of color channels convolved with pixel-level intensities of hematoxylin (H&E) and diaminobenzidine (DAB). The color components of these stains are significantly affected by laboratory conditions for sample preparation. Pathologists clearly observe variations in staining color, but DL models that strictly use color channels as inputs are largely undetectable. Therefore, trained models may perform poorly and be biased toward partial data sources.

[0026] Please refer to Figure 1, which shows a schematic diagram of a whole slide tissue imaging method provided in one embodiment of the present disclosure. In one embodiment, as shown in Figure 1, the whole slide tissue imaging method for assessing the expression level of HER2 in a target image includes the following steps:

[0027] S100 (step a): Acquire a target image / target tissue image, preferably whole slide imaging (WSI) immunohistochemical imaging.

[0028] S200 (step b): Obtain a dataset, enhance the staining components of the dataset, and use the enhanced dataset for associative learning to build a tumor segmentation model.

[0029] Notably, datasets obtained from different institutions are first subjected to a staining component augmentation (SC augmentation) operation. Using federated learning techniques, the augmented datasets are then used to construct tumor segmentation models. Federated learning is a machine learning technique that specifically involves training algorithms on multiple distributed edge devices or servers with local data samples. This approach differs significantly from traditional centralized machine learning approaches, which upload all local datasets to a single server, and more classical distributed approaches, which generally assume that local data samples are uniformly distributed. However, the disclosed federated learning approach allows multiple participants to build a common, robust machine learning model without sharing data, thereby addressing important issues such as data privacy, data security, data access rights, and heterogeneous data access. Therefore, using federated learning techniques, reliable models can be trained on data collected from users while ensuring the privacy of their data. The dataset includes multiple test images, which are immunohistochemical images.

[0030] Specifically, we use datasets obtained from National Taiwan University Hospital (NTUH) and the University of Warwick to perform local stain enhancement, upload the enhanced datasets to a federated learning server, and then build a tumor segmentation model using federated learning operations. The tumor segmentation model can automatically detect the location and boundary of tumors in medical images, segment tumor regions from normal regions, and quantify tumor size, volume, and characteristics, helping physicians quickly and accurately identify tumors in patients.

[0031] S300 (step c): Input the target image into a tumor segmentation model to determine the tumor region of the target image. Because the target image or target tissue image includes not only tumor regions but also normal regions, the purpose of this disclosure is to evaluate the expression level of human epidermal growth factor receptor 2 (HER2). Therefore, the tumor region within the target image or target tissue image must be determined in advance. Although this disclosure only clarifies the use of the method for evaluating the expression level of HER2, those skilled in the art can understand the necessary technical means from this disclosure and further apply this method to evaluate the expression levels of other genes or proteins. Therefore, using the method disclosed herein to evaluate the expression levels of other genes in the target tissue image should also be considered within the scope of this disclosure.

[0032] S400 (step d): Cell membrane staining detection is performed on the tumor area, and the tumor area is classified based on the cell membrane integrity and staining level, and a grade is given based on the classification results. Specifically, based on the staining results, the tumor area can be subdivided into negative, faint, weak, and strong. Cell membrane integrity is divided into complete and incomplete. By combining these, seven classification results can be obtained: negative, faint-incomplete, faint-complete, weak-incomplete, weak-complete, strong-incomplete, and strong-complete.

[0033] The disclosed whole-slide image processing method builds a tumor segmentation model using staining component expansion, improving compatibility with target images from different institutions and accurately representing tumor regions. Then, cell membrane staining detection is performed on the tumor region to classify the HER2 level in the target image. This method accurately evaluates the HER2 expression level in the target image and optimizes target images obtained from different institutions, enabling cross-institution model training without compromising user privacy.

[0034] See Figure 2, which shows another schematic diagram of a whole slide image processing method provided in one embodiment of the present disclosure. In one embodiment, the step of "acquiring a data set and enhancing staining components" further comprises the steps of:

[0035] S210: A test image in the dataset is acquired, and color separation is performed on the test image to acquire the color base and staining intensity of the dye.

[0036] Generally, stain component augmentation is an augmentation technique that expands the data distribution range during model training by simulating various histological images based on the source image. Due to differences in staining intensity and dye components between laboratories, stain component augmentation is necessary to minimize these differences. Ideally, pixels in an image are stained by only two dyes, such as H&E and DAB in IHC images. The stain component augmentation technique uses these two dyes to generate an enhanced image that better matches the distribution of histological data compared to typical RGB image enhancement. Specifically, it first acquires a test image in the dataset and performs color separation on the test image. For a given input IHC image, stain component augmentation calibrates its H&E and DAB channels through color separation. Image X is decomposed into a stain color base S and corresponding stain intensity A: T(X) = A*S, where T transforms the image into optical density space, and S consists of unit-length row vectors representing the H&E and DAB color components and an orthogonal background vector.

[0037] S220: Randomly scale and translate the staining intensity to obtain enhanced staining intensity;

[0038] Next, for each dye spot j and pixel i, the intensity A ij with a random coefficient α j Scaled by a random offset β j Shift with:A' ij =A ij *α j +β j .

[0039] S230: Randomly rotate the color base of the pigment to obtain an enhanced color base of the pigment.

[0040] Furthermore, the color component S of each dye spot j j is affected. (1, φ(S j ), θ(S j )) to S j If we use spherical coordinates, the angular displacement is: φ(S' j )=φ(S j )+δφ j ,θ(S' j )=θ(S j )+δθ j It is randomly rotated by the angle φ(S' j ) and θ(S' j ) is constrained between 0 and π / 2, so the perturbed color component S j remains in the valid optical density space. This perturbation randomly varies the color gamut of H&E and DAB around the original IHC image. All scaling α j , translation β j , rotation parameter δφ j and δθ j is randomly sampled from a uniform distribution with some predetermined range.

[0041] S240: The enhanced staining intensity and the enhanced color base are integrated to complete the enhancement of the staining components.

[0042] After integration, the enhanced image is: X'=T -1The resulting image is (A'*S'), completing the enhancement of the staining components. The effectiveness of the staining component augmentation technique can be evaluated using the Camelyon17 dataset. The Camelyon17 dataset contains thousands of high-resolution breast cancer tissue slice images from various cases, some of which contain cancerous regions and some of which contain normal tissue. Researchers can use these images to train and test deep learning models to automatically detect and identify cancerous regions, helping physicians make faster and more accurate diagnoses and treatment plans. This dataset demonstrates that the staining component augmentation technique in this application can fill distribution gaps in histological images and improve the stability and generalization of metastasis classification models trained using federated learning.

[0043] In one embodiment, the step of "using the enhanced datasets for federated learning to build a tumor segmentation model" further includes the steps of: pre-processing the multiple enhanced datasets and inputting the pre-processing results into a federated learning server to build a tumor segmentation model. The pre-processing is used to reduce the amount of data input into the federated learning server.

[0044] Specifically, preprocessing multiple enhanced datasets does not require uploading the original data (i.e., multiple enhanced datasets) to the federated learning server. Instead, mapping results and updates related to the original data are registered on the federated learning server. Compared to uploading the original data directly to the federated learning server, uploading the processed data can reduce the load on the federated learning server while maintaining the security of the original data, thereby avoiding the leakage of some private content contained in the original data.

[0045] In one embodiment, the step of "pre-processing the multiple enhanced datasets and inputting the pre-processing results into a federated learning server to construct a tumor segmentation model" further includes the steps of:

[0046] Step "Construct local and global models based on the learning rules of the federated learning server"

[0047] In the step "Training a local model using multiple enhanced datasets to obtain local model parameters," the multiple enhanced datasets can be obtained from different sources, such as the aforementioned NTUH and University of Warwick. The local model is locally trained based on the datasets to obtain local model parameters, such as weights, biases, etc., related to the original data.

[0048] Step "Input the local model parameters to the federated learning server and learn a global model through the clustering pipeline." In federated learning, a clustering pipeline refers to a set of processes and methods for clustering data in a distributed environment. The federated learning server inputs the local model parameters to the federated learning server, and then aggregates the local model parameters to obtain a global clustering model. Both the local model and the global model are based on federated learning rules.

[0049] The step "training a global model using multiple enhanced datasets and inputting the parameters of the trained global model into a federated learning server to construct a tumor segmentation model" is performed. Specifically, after the global model is trained in the federated learning server, the global model is distributed back to each local site. Each local site retrains the global model based on multiple enhanced datasets and inputs the trained parameters into the federated learning server. This learning process is repeated until the global model trained by the federated learning server meets certain pre-set requirements and is stopped. The trained global model becomes the required tumor segmentation model.

[0050] See Figure 3, which shows another schematic diagram of a whole slide image processing method provided in one embodiment of the present application. In one embodiment, step S400 (step d) "Performing cell membrane staining detection on tumor regions and classifying tumor regions based on cell membrane integrity and staining level" further includes the following steps:

[0051] S410 (Step d1): Detect nuclei in the tumor region and determine their locations. First, nuclei are detected in the IHC image. Scalar features for each pixel are calculated by combining hue, saturation, value, and RGB intensity variance. Based on these scalar features, a nuclear segmentation mask is generated using Otsu thresholding. The nuclear segmentation is then divided into individual nuclear segmentation masks by two successive partitioning processes, including connected component analysis, local peak detection, and a watershed algorithm. Uncovered nuclear components are further filtered based on their size and roundness.

[0052] S420 (step d2): Perform cell membrane detection based on the position of the nuclei to determine the position of the cell membrane corresponding to the nuclei. Specifically, the nuclear segmentation mask is dilated based on the position of the nuclei to determine the position of the cell membrane corresponding to the nuclei. For nuclei detected from the IHC image, first dilate the segmentation mask of each nucleus to determine its membrane area. The degree of dilation is dynamically calculated based on approximating the cell membrane as a hexagonal stack of circular objects.

[0053] S430 (step d3): Classify the cell membrane based on the staining level and obtain a pre-classification result. Specifically, color separation is performed on each pixel of the cell membrane and converted into an H&E-DAB representation; each pixel of the cell membrane is classified based on the staining intensity in the DAB channel according to a preset threshold, and a classification level for each pixel is obtained; and a pre-classification result is obtained based on the classification level for each pixel. Each nuclear membrane region is classified based on its staining level. Pixels in the membrane region are then converted into an H&E-DAB representation by color separation. Then, based on three preset thresholds for the DAB channel, they are divided into levels of negative, faint, weak, and strong, which is the pre-classification result. Each membrane is classified as the strongest staining level with enough pixels in the membrane region.

[0054] S440 (step d4): Classify the tumor region based on the pre-classification result and cell membrane integrity. Specifically, a skeletonization algorithm is used to obtain cell membrane integrity. The pre-classification result is refined based on cell membrane integrity to classify the tumor region. The cell membrane region is further classified based on staining integrity, resulting in classification results of negative, faint-incomplete, faint-complete, weak-incomplete, weak-complete, strong-incomplete, and strong-complete. For each membrane, the skeletonization algorithm is applied to pixels corresponding to the staining level of the membrane region and one level weaker staining level. The skeleton contour is then processed by a closing operation. If the skeleton contour is closed and encloses the main area of the membrane, the membrane is classified as complete; otherwise, it is considered incomplete.

[0055] In one embodiment, the classification results are graded according to preset criteria, including the ASCO CAP 2018 guidelines. The whole slide image processing method includes determining the HER2 level in the target image based on the ASCO CAP 2018 guidelines and the classification results. Specifically, the American Society of Clinical Oncology (ASCO) and the College of American Pathologists (CAP) jointly published guidelines for cancer diagnosis and treatment in 2018, providing physicians and pathologists with the latest diagnostic and treatment recommendations. The obtained classification results can be used to determine the HER2 level in the target image as 0, 1+, 2+, or 3+ according to the guidelines.

[0056] Based on the same technical concept, the present disclosure also provides an embodiment of a tissue image processing system for whole slide imaging, including a memory, a processor, and a computer program stored on the memory, wherein the processor executes the computer program to perform steps of the whole slide image processing method in the embodiment.

[0057] The disclosed whole slide image processing system builds a tumor segmentation model using staining component expansion to enhance compatibility with target images from different institutions and accurately display tumor regions. Then, cell membrane staining detection is performed on the tumor region to classify the HER2 level of the target image. This system accurately evaluates the HER2 level in the target image and optimizes target images obtained from different institutions, enabling cross-institution model training without compromising user privacy.

[0058] Based on the same technical concept, the present disclosure further provides a computer program product (e.g., a computer-readable recording medium) including a computer program / instruction, which, when executed by a processor, performs the steps of the hole slide image processing method in the embodiment.

[0059] In the above embodiments, the description of each embodiment focuses on different aspects. For parts that are not described or recorded in detail in one embodiment, reference can be made to the relevant descriptions of other embodiments. In addition, the above embodiments can be freely combined as needed.

[0060] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to Provisional Application No. 63 / 550,591, filed February 6, 2024, the contents of which are incorporated herein by reference in their entirety.

Claims

1. (a) acquiring a target image; (b) obtaining a dataset, enhancing staining components of the dataset, and using the enhanced dataset for associative learning to construct a tumor segmentation model; the dataset includes test images, the test images being immunohistochemical images; (c) inputting the target image into the tumor segmentation model to determine a tumor region in the target image; (d) performing cell membrane staining detection on the tumor region and further classifying the tumor region based on the cell membrane integrity and the cell membrane staining level; (e) Ranking based on the classification results; Tissue imaging processing methods.

2. Obtaining the data set and enhancing staining components of the data set includes: obtaining the test image of the dataset, performing color separation on the test image, and obtaining color bases and staining intensities of dyes in the test image; Randomly scaling and randomly translating the staining intensity of the dye to enhance the staining intensity of the dye; randomly rotating the color base of the dye to obtain an enhanced color base of the dye; Integrating the enhanced color base of the pigment and the enhanced staining intensity to complete the enhancement of the staining components. further comprising:

10. The tissue imaging process of claim 1.

3. using the enhanced dataset for federated learning to construct the tumor segmentation model, pre-processing the enhanced dataset and inputting the pre-processing results into a federated learning server to construct the tumor segmentation model; further comprising:

3. The tissue imaging processing method according to claim 1 or 2.

4. Preprocessing the enhanced dataset and inputting a preprocessing result into a federated learning server to construct the tumor segmentation model includes: constructing a local model and a global model based on the learning rules of the federated learning server; training a local model on the enhanced dataset to obtain a plurality of parameters of the local model; inputting the parameters of the local models into the federated learning server and training the global model through a clustering pipeline; training the global model on the enhanced dataset, and inputting parameters of the trained global model into the federated learning server to construct the tumor segmentation model; further comprising:

4. The tissue imaging process of claim 3.

5. (d) is (d1) performing nuclei detection on the tumor region to determine the location of nuclei in the tumor region; (d2) performing cell membrane detection based on the location of the nuclei to determine the location of the cell membrane corresponding to the nuclei; (d3) classifying the cell membrane based on the staining level of the cell membrane to obtain a pre-classification result; (d4) classifying the tumor region based on the pre-classification results and the cell membrane integrity; further comprising:

10. The tissue imaging process of claim 1.

6. (d2) is Expanding the segmentation mask of the nucleus based on the position of the nucleus, and determining the position of the cell membrane corresponding to the nucleus. further comprising:

6. The tissue imaging process of claim 5.

7. (d3) is performing color separation on each pixel of the cell membrane and converting it into a representation in hematoxylin-diaminobenzidine; Classifying each pixel of the cell membrane according to a preset threshold based on the staining intensity of the diaminobenzidine channel to obtain a classification level for each pixel; obtaining a pre-classification result based on the classification level of each pixel; further comprising:

10. The tissue imaging process of claim 1.

8. (d4) is obtaining the integrity of the cell membrane using a skeletonization algorithm; refining the pre-classification results based on the integrity of the cell membrane to classify the tumor region; further comprising:

6. The tissue imaging process of claim 5.

9. Memory and a processor; a computer program stored in the memory; and Including, The processor executes the computer program to implement the tissue imaging processing method according to any one of claims 1 to 8. Tissue imaging processing system.

10. Contains computer programs / instructions, The computer program / instructions, when executed by a processor, perform the tissue imaging processing method of any one of claims 1 to 8. A computer-readable medium for tissue imaging processing.

Images