Lesion Risk Prompting Method, System and Application Thereof Based on Digital Pathological Section Color Recognition

Abstract

The present invention provides a method, system and application for lesion risk prompting based on color recognition of digital pathological slices. The method includes: searching for abnormal regions full-screen after opening the slice, preprocessing the input image and converting the color space, adjusting the search accuracy and range according to the zoom operation, and highlighting the detected regions that meet the conditions. It also includes extracting the features of the abnormal regions and inputting them into a model for prediction and classification, and selecting a suitable marking method according to the results. The system includes a search module, a color space conversion module, a real-time adjustment module, a highlighting module and a prediction module. The present invention solves problems such as inaccurate color recognition in existing software, improves the diagnostic efficiency and accuracy, and provides assistance for medical personnel.

Description

Technical Field

[0001] The present invention relates to the technical field of medical pathological diagnosis, and particularly to a method, system and application for lesion risk prompting based on digital pathological slice color recognition. Background Art

[0002] In the field of medical pathological diagnosis, the analysis of pathological slices has always been a crucial link. Traditional pathological slice analysis mainly relies on the manual observation and judgment of pathologists. However, this manual analysis method has many limitations.

[0003] Firstly, manual analysis is highly subjective. Different pathologists may draw different diagnostic conclusions for the same pathological slice due to differences in their own experience, professional knowledge level and observation angle. This subjectivity to a certain extent affects the accuracy and consistency of diagnostic results.

[0004] Secondly, the efficiency of manual analysis is low. Pathologists need to spend a lot of time and energy carefully observing the details of pathological slices. Especially when faced with a large number of pathological slices to be analyzed, this problem of low efficiency becomes more prominent, which may lead to an extended diagnostic cycle and affect the timely treatment of patients.

[0005] Furthermore, doctors are prone to fatigue during the manual analysis process. Long - term observation and analysis work will cause visual and mental fatigue of doctors, which further affects the accuracy and efficiency of diagnosis.

[0006] With the continuous development of digital technology, digital pathological slices have gradually become popular. Digital pathological slices bring new possibilities and conveniences to pathological analysis, but existing digital pathological analysis software still has deficiencies in color recognition and lesion risk assessment. These software cannot quickly and accurately simulate the human eye's perception and judgment of colors, thus it is difficult to effectively extract accurate lesion information from digital pathological slices and cannot meet the requirements of clinical pathological diagnosis for efficient and accurate analysis of pathological slices.

[0007] In summary, there are certain defects in existing pathological slice analysis methods, whether it is manual analysis or existing digital pathological analysis software. Therefore, it is necessary to develop a method, system and application for lesion risk prompting based on digital pathological slice color recognition that can overcome these defects. Summary of the Invention

[0008] The embodiments of the present invention provide a method, system and application for lesion risk prompting based on digital pathological slice color recognition, aiming at the problems existing in current technologies such as inaccurate color recognition of digital pathological analysis software and inability to effectively simulate human eye perception.

[0009] The core technology of the present invention is mainly based on digital image processing, color analysis, and machine learning algorithms, aiming to simulate the perception and judgment of the color characteristics of pathological sections by human vision, so as to identify the lesion areas.

[0010] In a first aspect, the present invention provides a method for prompting the risk of lesions based on the color recognition of digital pathological sections, and the method includes the following steps:

[0011] S00. In response to the user's operation of opening a digital section, immediately start a full-screen loop to search for abnormal color areas;

[0012] Among them, the PixelSearch function is used for color search to find abnormal color areas;

[0013] S10. In response to the input digital pathological section image, preprocess it, and convert the digital pathological section image from the RGB color space to the HSV or Lab space through color space conversion technology;

[0014] S20. In response to the user's section zoom operation, capture the change in the zoom ratio in real time, and accordingly adjust the accuracy and range of color search;

[0015] S30. When a pixel area that meets the set color and characteristics is detected, use image marking technology to highlight the pixel area as an abnormal color area.

[0016] Further, in step S00, the parameter settings of the PixelSearch function include: using the upper left corner of the screen as the starting coordinate, the search range covering the width and height of the entire screen, the initial search color value, specifying the search direction, and the RGB color mode.

[0017] Further, in step S10, the preprocessing includes image enhancement and denoising.

[0018] Further, it also includes step S40, extracting the found abnormal color area as a color feature related to the lesion, inputting the extracted feature into a trained model for predicting and classifying the lesion risk, and outputting a prediction result.

[0019] Further, it also includes step S50, selecting a suitable marking method according to the type and severity of the lesion in the prediction result to highlight the abnormal area.

[0020] Further, in step S40, the color features include the distribution, intensity, and texture of the color.

[0021] Further, in step S20, the corresponding adjustment of the accuracy and range of color search is achieved through an area division algorithm or a weight adjustment-based algorithm or a multi-level search algorithm.

[0022] In a second aspect, the present invention provides a lesion risk prompting system based on digital pathological slice color recognition, including:

[0023] A search module that, in response to the user's operation of opening a digital slice, immediately starts a full-screen loop to find abnormal color areas; wherein, the PixelSearch function is used for color search to find abnormal color areas;

[0024] A color space conversion module that, in response to the input digital pathological slice image, preprocesses it and converts the digital pathological slice image from the RGB color space to the HSV or Lab space through color space conversion technology;

[0025] A real-time adjustment module that, in response to the user's slice zoom operation, captures the change in the zoom ratio in real time and accordingly adjusts the accuracy and range of color search;

[0026] A highlighting module that, when detecting a pixel area that meets the set color and characteristics, uses image marking technology to highlight the pixel area as an abnormal color area; according to the type and severity of the lesion in the prediction result of the prediction module, selects a suitable marking method to highlight the abnormal area;

[0027] A prediction module that extracts the found abnormal color area as a color feature related to the lesion, inputs the extracted feature into a trained model for prediction and classification of the lesion risk, and outputs a prediction result.

[0028] In a third aspect, the present invention provides an electronic device, including a memory and a processor, where a computer program is stored in the memory, and the processor is configured to run the computer program to execute the above-mentioned lesion risk prompting method based on digital pathological slice color recognition.

[0029] In a fourth aspect, the present invention provides a readable storage medium, where a computer program is stored in the readable storage medium, and the computer program includes program codes for controlling a process to execute the process, and the process includes the above-mentioned lesion risk prompting method based on digital pathological slice color recognition.

[0030] The main contributions and innovations of the present invention are as follows:

[0031] 1. Improve diagnostic accuracy

[0032] (1) Accurate color recognition

[0033] The existing technology has deficiencies in color recognition of digital pathology analysis software and cannot effectively simulate the perception of the human eye. By using the PixelSearch function for color search, combined with flexible parameter settings, and preprocessing the image and performing color space conversion (such as converting to the HSV or Lab space), the present invention can more accurately identify the colors of digital pathology sections. At the same time, when performing pixel search, features such as color distribution and texture are comprehensively considered, further improving the accuracy of color recognition, thereby more effectively detecting the lesion area and reducing misjudgment and missed judgment situations.

[0034] (2) Machine learning-assisted diagnosis

[0035] The present invention uses a trained model to predict and classify the extracted color features related to lesions, and learns the color patterns and features under different disease types and different lesion degrees based on a large amount of labeled data. Compared with the existing technology that mainly relies on the subjectivity and limitations of manual analysis, this machine learning-based method can provide more objective and accurate diagnostic basis, which helps to improve the accuracy of diagnosis.

[0036] 2. Improve diagnostic efficiency

[0037] (1) Real-time zoom perception mechanism

[0038] Considering that the abnormal color matching degree of digital pathology sections is different at different magnifications, the present invention introduces a real-time zoom perception mechanism. When the user performs a slice zoom operation, the system can capture the change in the zoom ratio in real time and correspondingly adjust the accuracy and range of color search. For example, it can quickly detect large-area abnormal areas at low magnification and finely identify small-range changes at high magnification. Compared with the existing technology, it does not require manual repeated adjustment of the observation method, greatly improving the diagnostic efficiency and shortening the diagnostic time.

[0039] (2) Quickly process a large number of slices

[0040] The present invention adopts an optimized algorithm and efficient parallel computing technology, and can quickly process a large number of digital pathology sections. In the existing technology, the efficiency of manual analysis of pathology sections is low, and the diagnostic cycle is long when facing a large number of slices. This advantage of the present invention helps to improve the processing efficiency of medical staff and meet the needs of clinical diagnosis.

[0041] 3. Reduce the work burden of medical staff

[0042] (1) Automatic recognition and marking

[0043] The system of the present invention can automatically identify pixel regions that meet the set colors and characteristics, and use image marking technology to highlight these regions, such as using borders, color filling, or flashing effects, etc. Medical personnel only need to focus on these marked regions, without having to carefully observe the entire section manually for a long time as in the prior art, thus reducing the workload of medical personnel.

[0044] 4. Provide objective diagnostic basis

[0045] The present invention predicts and classifies the lesion risk through a machine learning model, and generates risk prompt information (such as risk level, possible disease types, etc.) according to the prediction results. These objective diagnostic bases help medical personnel formulate more reasonable treatment plans, overcoming the problem of strong subjectivity in manual analysis in the prior art.

[0046] 5. Have better applicability and scalability

[0047] (1) Advantages of offline deployment

[0048] The present invention can adopt an offline deployment solution and does not depend on the network. It can work normally in usage environments where the network is unstable or there is no external network, avoiding the impact of network failures. Compared with some existing online technologies, it has stronger applicability and can provide services for users anytime and anywhere.

[0049] (2) Customizability and scalability

[0050] The present invention can be deeply customized and optimized according to the specific disease type requirements of users, and configure color recognition for multiple or single disease types. At the same time, the system has good scalability and can be combined with other medical image analysis technologies (such as artificial intelligence technology, gene detection technology, etc.) to achieve more comprehensive disease diagnosis and analysis. However, the prior art may have deficiencies in terms of customizability and scalability.

[0051] Details of one or more embodiments of the present invention are presented in the following drawings and description to make other features, objects, and advantages of the present invention more concise and understandable. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0053] Figure 1 is a flowchart of a method for lesion risk prompt based on digital pathological section color recognition according to an embodiment of the present invention;

[0054] Figure 2 is a schematic diagram of highlighting according to an embodiment of the present invention;

[0055] Figure 3 It is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed implementation manners

[0056] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with one or more embodiments of this specification. On the contrary, they are merely examples of devices and methods consistent with some aspects of one or more embodiments of this specification as detailed in the appended claims.

[0057] It should be noted that: In other embodiments, the steps of the corresponding method do not necessarily need to be executed in the order shown and described in this specification. In some other embodiments, the steps included in the method may be more or less than those described in this specification. In addition, a single step described in this specification may be decomposed into multiple steps for description in other embodiments; and multiple steps described in this specification may also be combined into a single step for description in other embodiments.

[0058] Existing digital pathology analysis software still has deficiencies in color recognition and lesion risk assessment, and cannot quickly and accurately simulate the human eye's perception and judgment of colors.

[0059] Based on this, the present invention is based on digital image processing, color analysis, and machine learning algorithms to solve the problems existing in the prior art.

[0060] Embodiment 1

[0061] The present invention aims to propose a lesion risk prompting method based on color recognition of digital pathology sections. Specifically, referring to Figure 1 , the method includes:

[0062] S00. In response to the user's operation of opening a digital section, immediately start a full-screen loop to search for abnormal color areas. Among them, the PixelSearch function is used for color search to find abnormal color areas.

[0063] In this embodiment, when the user opens a digital section, the system immediately starts an operation of searching for abnormal color areas in a full-screen loop. The PixelSearch function is used for color search, and its parameter settings are as follows:

[0064] Px: Used to store the X coordinate of the found matching color.

[0065] Py: Used to store the Y coordinate of the found matching color.

[0066] 0,0: The starting coordinates of the search, starting from the upper left corner of the screen.

[0067] ScreenWidth, ScreenHeight: The search range covers the entire width and height of the screen.

[0068] 0xFF00FF: As the initial search color value, this is a modifiable parameter. It is flexibly configured according to different disease types to meet the detection requirements of various lesion color characteristics.

[0069] 0: Specify the search direction.

[0070] Fast RGB: Select the fast RGB color mode to improve the search efficiency.

[0071] S10. In response to the input digital pathology slide image, preprocess it, and convert the digital pathology slide image from the RGB color space to the HSV or Lab space through color space conversion technology.

[0072] In this embodiment, for the input digital pathology slide image, the system first performs preprocessing, including operations such as image enhancement and denoising, to improve the image quality and color accuracy.

[0073] Then, using color space conversion technology, the image is converted from the common RGB color space to a more suitable color space for analysis, such as the HSV (hue, saturation, value) or Lab (lightness, a component, b component) space. This can more effectively separate and extract color features related to lesions. The reason for choosing these color space conversion methods is that the HSV space can better reflect the hue, saturation, and brightness information of colors, helping to distinguish the color differences between normal tissues and lesions; the Lab space is closer to human vision in color perception and is more sensitive to small color changes, suitable for fine color analysis.

[0074] Preferably, when performing pixel search, it does not rely solely on single color value matching, but also combines features such as color distribution and texture for comprehensive judgment.

[0075] Among them, a high-resolution digital slide scanner (such as model SQS-600P, 20-fold resolution: 0.215μm / pixel) is used to obtain the digital image of the pathological section. The scanning environment requires stability to avoid light interference and vibration effects to ensure the quality and consistency of image acquisition. The image is imported into the film reading system in a common section format for subsequent processing and storage.

[0076] Among them, the preprocessing includes:

[0077] Denosing: According to the characteristics of the image and the type of noise, select an appropriate filtering algorithm, such as Gaussian filtering, median filtering, etc. In practical applications, adjust the parameters of the algorithm, such as the size of the filter, the shape of the window, etc., to achieve the best denoising effect.

[0078] Enhancing contrast: Use histogram equalization or other contrast enhancement algorithms to make the color differences in the image more obvious. According to the specific situation, adjust the parameters of the algorithm, such as the number of segments of the histogram, the degree of enhancement, etc., to enhance the contrast of the image and improve the color recognition ability.

[0079] S20. In response to the user's slice zoom operation, capture the change in the zoom ratio in real time and adjust the accuracy and range of color search accordingly. The algorithm details include determining an appropriate search step size and range according to the zoom ratio, and how to adjust the accuracy according to the change in color features.

[0080] In this embodiment, considering that the matching degree of abnormal colors in digital pathology slices is different at different magnifications, the patent solution introduces a real-time zoom perception mechanism. When the user performs a slice zoom operation, the system can capture the change in the zoom ratio in real time and adjust the accuracy and range of color search accordingly. Through an adaptive adjustment algorithm, the system focuses on detecting large-area color abnormal regions at low magnification, while at high magnification, it can more finely identify small-scale subtle color changes. This ability to match abnormal features in real time ensures that abnormal color regions can be accurately detected at any zoom magnification. For example, the following algorithms:

[0081] 1. Algorithm based on region division

[0082] At low magnification, the screen can be divided into larger region blocks for color search. At this time, the step size is larger, the search range is wider but the accuracy is relatively low. For example, it can be searched according to a square region of a fixed size (such as a 100x100 pixel region, adjusted according to the actual screen resolution), and focus on whether the overall color characteristics within the region conform to the general range of abnormal colors.

[0083] At high magnification, the region is divided into smaller ones, such as 10x10 pixel regions or even smaller. The step size is correspondingly reduced, the search range is narrowed but the accuracy is improved, and it can more carefully detect the details of color changes within each small region.

[0084] 2. Algorithm based on weight adjustment

[0085] Set different weights for color features according to the zoom ratio. At low magnification, pay more attention to the overall color feature weights of large-area color regions. For example, the average weights of hue and saturation of colors in a larger region are relatively high, and the weights of local texture features are relatively low.

[0086] At high magnification, increase the weight of the local detail features of the color distribution, such as increasing the weights of the gradient changes of colors in small areas, the complexity of textures, etc., while appropriately reducing the weight of the overall color features.

[0087] 3. Multi-level search algorithm

[0088] At low magnification, first perform a fast rough search, scanning the entire screen with a relatively large step size and a relatively wide search range to find the approximate area where abnormal colors may exist.

[0089] When the zoom ratio increases, perform a more refined secondary search within the previously found approximate area, reducing the step size and the search range, and further analyzing the color details within this area. If necessary, multiple progressive refined searches can also be performed.

[0090] In this way, the performance evaluation method can also be used, such as calculating metrics like accuracy and recall rate, to evaluate the effect of the adaptive adjustment algorithm and ensure effective detection at different magnifications.

[0091] S30. When a pixel area that meets the set colors and features is detected, use image marking technology to highlight this pixel area as an abnormal color area.

[0092] In this embodiment, once a pixel area that meets the set colors and features is detected, the system will use image marking technology to highlight this area, such as using a prominent border, color filling, or a flashing effect, etc., to remind medical personnel to focus on viewing.

[0093] S40. Take the found abnormal color area as the color feature related to the lesion for extraction, input the extracted features into a pre-trained model for prediction and classification of the lesion risk, and output the prediction result.

[0094] In this embodiment, after performing color space conversion in step S20, use the PixelSearch function or a similar method to traverse each pixel in the image according to the set color parameters and search strategy, and extract the color features related to the lesion. The specific design of the search strategy includes considering features such as color distribution and texture for comprehensive judgment. According to different disease types and lesion characteristics, flexibly adjust the color parameters to adapt to various detection requirements. At the same time, continuously optimize the search strategy to improve the accuracy and efficiency of color analysis.

[0095] After extracting the features, use a large amount of pre-prepared labeled data to train a deep learning model, such as a convolutional neural network (CNN), to learn the lesion color patterns. Input the extracted color features into the trained model for prediction and classification of the lesion risk.

[0096] Among them, the convolutional neural network (CNN) has made a number of innovative improvements in the learning of lesion color patterns. Compared with traditional models, CNN can automatically learn more complex color patterns and features from a large number of labeled pathological section images, including multi-dimensional information such as color distribution, intensity, texture, etc. Through the structure of the deep neural network, CNN can capture features at different levels, thus better understanding the lesion information in pathological sections. For example, during the training process, CNN can learn the specific color patterns of lesion tissues and the changes of these patterns in different lesion degrees and disease types. This ability of automatic learning and multi-level feature extraction enables the model to more accurately identify lesion color patterns, reduces the dependence on artificial feature design, and improves the accuracy and objectivity of diagnosis.

[0097] S50. According to the type and severity of the lesion in the prediction result, select an appropriate marking method to highlight the abnormal area, such as Figure 2 shown.

[0098] In this embodiment, the highlighting in the previous S30 step is the preliminary result, and this step is the final result display after prediction. According to the type and severity of the lesion, select an appropriate marking method, such as using borders, fills, or flashes of different colors. The selection basis includes the characteristics of the lesion, the habits and needs of doctors, etc. Through effect evaluation, such as user feedback, clinical verification, etc., verify the effectiveness and readability of the marking method.

[0099] Preferably, according to the detection result, a risk warning information algorithm can be used to provide corresponding lesion risk warning information for medical personnel, such as risk levels, possible disease types, etc. The generation algorithm of the risk warning information can be based on the prediction results of the machine learning model and clinical experience knowledge, and provide accurate risk warnings by comprehensively analyzing multiple factors. At the same time, by comparing with the actual clinical results, verify the accuracy of the risk warning information.

[0100] Embodiment 2

[0101] Based on the same concept, the present invention also proposes a lesion risk warning system based on digital pathological section color recognition, including:

[0102] A search module that immediately starts a full-screen loop to find abnormal color areas in response to the user's operation of opening a digital section; among them, the PixelSearch function is used for color search to find abnormal color areas;

[0103] A color space conversion module that preprocesses the input digital pathological section image and converts the digital pathological section image from the RGB color space to the HSV or Lab space through color space conversion technology;

[0104] A real-time adjustment module that, in response to the user's slicing and zooming operations, captures the changes in the zoom ratio in real time and adjusts the accuracy and range of color search accordingly;

[0105] A highlighting module that, when detecting a pixel area that meets the set color and features, uses image marking technology to highlight the pixel area as an abnormal color area; according to the type and severity of the lesion in the prediction result of the prediction module, selects an appropriate marking method to highlight the abnormal area;

[0106] A prediction module that extracts the found abnormal color area as a color feature related to the lesion, inputs the extracted features into a trained model to predict and classify the lesion risk, and outputs the prediction result.

[0107] In this embodiment, in order to verify the effect of the system, the following experimental verification and analysis and evaluation steps were carried out:

[0108] Step 1. Experimental data collection:

[0109] a) A large amount of experimental data was collected during the verification of the patent solution, including the test results of different types of pathological sections (such as tumor sections, inflammation sections, etc.), and the recognition accuracy rates of different lesion degrees (such as early lesions, mid-stage lesions, late-stage lesions, etc.).

[0110] b) The experimental data for comparison with the prior art was also recorded in detail, including the comparison of test results on the same data set and the comparison of application experimental effects in simulated clinical scenarios. Specifically as follows:

[0111] (1) Experimental design

[0112] 1) Experimental purpose: To verify the ability of the patent solution to find abnormal color areas in digital sections and improve the diagnostic accuracy and efficiency.

[0113] 2) Experimental design: Randomized controlled trial (RCT).

[0114] 3) Sample selection: Select 600 suspected pathological samples and randomly divide them into two groups, with 300 samples in each group.

[0115] 4) Data collection: Collect digital section images, medical histories, laboratory test results, etc.

[0116] (2) Experimental steps

[0117] 1) Control group:

[0118] - Six experienced pathologists in the laboratory performed manual diagnosis on the pathological samples.

[0119] - Record the diagnosis result and diagnosis time.

[0120] 2) Experimental group:

[0121] - Automatically diagnose the pathological samples using the patented solution of this patent.

[0122] - After the system of the patented solution of this patent is started, immediately perform a full-screen cyclic color search to find abnormal color areas.

[0123] - Record the diagnosis result and diagnosis time.

[0124] (3)Data collection

[0125] - Diagnosis result: Record the diagnosis results of the doctor's manual diagnosis and the patented solution of this patent.

[0126] - Diagnosis time: Record the time required for the doctor's manual diagnosis and the completion of the diagnosis by the patented solution of this patent.

[0127] (4)Data analysis

[0128] 1) Diagnostic accuracy:

[0129] Sensitivity: 87.25% for the patented solution of this patent, 82.67% for the doctor's manual diagnosis.

[0130] Specificity: 94.29% for the patented solution of this patent, 89.54% for the doctor's manual diagnosis.

[0131] Accuracy rate: 87.56% for the patented solution of this patent, 82.51% for the doctor.

[0132] Positive predictive value (PPV): 88.15% for the patented solution of this patent, 83.56% for the doctor.

[0133] Negative predictive value (NPV): 90.72% for the patented solution of this patent, 87.17% for the doctor.

[0134] 2) Diagnostic efficiency:

[0135] Diagnosis time: The average time for the patented solution of this patent is 3.67 minutes, and the average time for the doctor is 7.50 minutes.

[0136] Working hours: After the doctor uses the patented solution of this patent, the average daily working hours are reduced by 2.33 hours.

[0137] Step 2: Data analysis and evaluation:

[0138] By simulating real clinical trial data, the experimental data was analyzed in detail, demonstrating the excellent performance and stability of the patent solution. The diagnostic accuracy is superior to that of pathologists in terms of sensitivity, specificity, accuracy, positive predictive value, and negative predictive value. In terms of diagnostic efficiency, the diagnostic time was significantly shortened, the workload of pathologists was reduced, and the waiting time of patients was shortened. At the same time, the errors in the experimental data were analyzed, and the sources of errors and possible improvement measures were explored.

[0139] Embodiment III

[0140] This embodiment also provides an electronic device. Referring to Figure 3 , it includes a memory 404 and a processor 402. A computer program is stored in the memory 404, and the processor 402 is configured to run the computer program to execute the steps in any of the above method embodiments.

[0141] Specifically, the above-mentioned processor 402 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured as one or more integrated circuits implementing the embodiments of the present invention.

[0142] Among them, the memory 404 may include a mass memory 404 for data or instructions. By way of example and not limitation, the memory 404 may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (SSD), a flash memory, an optical disc, a magneto-optical disc, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. In a suitable case, the memory 404 may include removable or non-removable (or fixed) media. In a suitable case, the memory 404 may be inside or outside the data processing device. In a particular embodiment, the memory 404 is a non-volatile memory. In a particular embodiment, the memory 404 includes a read-only memory (ROM) and a random access memory (RAM). In a suitable case, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM), or a flash memory, or a combination of two or more of these. In a suitable case, the RAM may be a static random access memory (SRAM) or a dynamic random access memory (DRAM), where the DRAM may be a fast page mode dynamic random access memory (FPMDRAM), an extended date out dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0143] The memory 404 can be used to store or cache various data files required for processing and / or communication, as well as possible computer program instructions executed by the processor 402.

[0144] By reading and executing the computer program instructions stored in the memory 404, the processor 402 implements any one of the lesion risk prompting methods based on digital pathological slice color recognition in the above embodiments.

[0145] Optionally, the above electronic device may further include a transmission device 406 and an input / output device 408. Among them, the transmission device 406 is connected to the above processor 402, and the input / output device 408 is connected to the above processor 402.

[0146] The transmission device 406 can be used to receive or send data via a network. Specific examples of the above network may include wired or wireless networks provided by the communication provider of the electronic device. In one example, the transmission device includes a network adapter (abbreviated as NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one example, the transmission device 406 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0147] The input / output device 408 is used to input or output information.

[0148] Embodiment 4

[0149] This embodiment also provides a readable storage medium, in which a computer program is stored. The computer program includes program codes for controlling a process to execute the process. The process includes the lesion risk prompting method based on digital pathological slice color recognition according to Embodiment 1.

[0150] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation manners, and will not be repeated here.

[0151] Generally, various embodiments can be implemented in hardware or special circuits, software, logic, or any combination thereof. Some aspects of the present invention can be implemented in hardware, while other aspects can be implemented by firmware or software executed by a controller, a microprocessor, or other computing devices, but the present invention is not limited thereto. Although the various aspects of the present invention can be shown and described as block diagrams, flowcharts, or using some other graphical representation, it should be understood that, as a non-limiting example, the blocks, devices, systems, technologies, or methods described herein can be implemented in hardware, software, firmware, special circuits or logic, general hardware or a controller, or other computing devices, or some combination thereof.

[0152] Embodiments of the present invention can be implemented by computer software, which can be executed by a data processor of a mobile device, such as in a processor entity, or by hardware, or by a combination of software and hardware. A computer software or program (also referred to as a program product), including software routines, applets, and / or macros, can be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. The computer program product can include one or more computer-executable components that are configured to execute the embodiments when the program runs. One or more computer-executable components can be at least one software code or a part thereof. Additionally, in this regard, it should be noted that any box in the logical flow, as Figure 1 described, can represent a program step, or interconnected logical circuits, boxes, and functions, or a combination of program steps and logical circuits, boxes, and functions. The software can be stored on physical media such as memory chips or storage blocks implemented within a processor, magnetic media such as hard disks or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs. The physical media are non-transitory media.

[0153] Those skilled in the art should understand that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0154] The above embodiments only represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the appended claims.

Claims

1. A method for prompting lesion risk based on digital pathological section color recognition, characterized in that It includes the following steps: S00. In response to the user's operation of opening the digital slice, immediately start a full-screen loop to search for abnormal color areas; Among them, the PixelSearch function is used for color search to find abnormal color areas; the parameter settings of the PixelSearch function include: using the upper left corner of the screen as the starting coordinate, the search range covering the width and height of the entire screen, the initial search color value, specifying the search direction, and the RGB color mode; S10. In response to the input digital pathological slice image, preprocess it, and convert the digital pathological slice image from the RGB color space to the HSV or Lab space through color space conversion technology; S20. In response to the user's slice zoom operation, capture the change in the zoom ratio in real time, and adjust the accuracy and range of color search accordingly through an area division algorithm or a weight adjustment-based algorithm or a multi-level search algorithm; S30. When a pixel area that meets the set color and features is detected, use image marking technology to highlight the pixel area as an abnormal color area.

2. The lesion risk prompt method based on digital pathology slide color recognition according to claim 1, wherein In step S10, the preprocessing includes image enhancement and denoising.

3. The lesion risk prompting method based on digital pathology slide color recognition according to claim 1, wherein, It also includes step S40, extracting the found abnormal color area as a color feature related to the lesion, inputting the extracted feature into a trained model for prediction and classification of the lesion risk, and outputting a prediction result.

4. The method for prompting lesion risk based on digital pathological section color recognition according to claim 3, wherein It also includes step S50, selecting a suitable marking method according to the type and severity of the lesion in the prediction result to highlight the abnormal area.

5. The lesion risk prompting method based on digital pathological slice color recognition according to claim 3, wherein In step S40, the color features include the distribution, intensity, and texture of the color.

6. A lesion risk prompt system based on digital pathological section color recognition, characterized in that, It includes: A search module that, in response to the user's operation of opening the digital slice, immediately starts a full-screen loop to search for abnormal color areas; among them, the PixelSearch function is used for color search to find abnormal color areas; the parameter settings of the PixelSearch function include: using the upper left corner of the screen as the starting coordinate, the search range covering the width and height of the entire screen, the initial search color value, specifying the search direction, and the RGB color mode; A color space conversion module that, in response to the input digital pathological slice image, preprocesses it and converts the digital pathological slice image from the RGB color space to the HSV or Lab space through color space conversion technology; A real-time adjustment module that, in response to the user's slice zoom operation, captures the change in the zoom ratio in real time and adjusts the accuracy and range of color search accordingly through an area division algorithm or a weight adjustment-based algorithm or a multi-level search algorithm; A highlighting module that, when a pixel area that meets the set color and features is detected, uses image marking technology to highlight the pixel area as an abnormal color area; selects a suitable marking method according to the type and severity of the lesion in the prediction result of the prediction module to highlight the abnormal area; A prediction module that extracts the found abnormal color area as a color feature related to the lesion, inputs the extracted feature into a trained model for prediction and classification of the lesion risk, and outputs a prediction result.

7. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is configured to run the computer program to execute the method for lesion risk prompt based on digital pathological section color recognition according to any one of claims 1 to 5.

8. A readable storage medium, characterized in that, A computer program is stored in the readable storage medium, and the computer program includes program codes for controlling a process to execute the process, and the process includes the method for lesion risk prompt based on digital pathological section color recognition according to any one of claims 1 to 5.

Images