Medical image key area detection system and method

Abstract

The invention relates to the technical field of medical image area detection, and discloses a medical image key area detection system and method.The medical image key area detection system comprises an image obtaining module, an image preprocessing module, a key area positioning module, a feature extraction module and a classification and recognition module. According to the key area detection system of the medical image, the key anatomical structure or the lesion area in the image can be automatically identified and positioned through the system, so that the diagnosis efficiency of a doctor is greatly improved, an important reference can be provided for clinical decision-making of the doctor, and the accuracy of clinical decision-making is improved by accurately identifying the lesion area. According to the technical scheme, doctors can more accurately evaluate the severity of the illness state, formulate a more effective treatment scheme and even predict the development trend of the illness, so that the method has important significance in improving the treatment effect and the survival rate of patients, meanwhile, errors caused by human factors are reduced through automatic analysis, and the diagnosis reliability is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image region detection, and particularly to a key region detection system and method for medical images. Background Art

[0002] Medical images are images that reflect the internal structure or internal function of the human anatomical region. Medical images are composed of a set of image elements (pixels or voxels), which are discrete image representations generated through sampling or reconstruction, and can map numerical values to different spatial positions. These images can reflect the internal structure or internal function of the anatomical region and are an important means for modern medical diagnosis and treatment. Medical images mainly consist of parts such as pixel depth, photometric interpretation, metadata, and pixel data.

[0003] The processing and analysis of medical images are important links in medical imaging technology, mainly including steps such as image enhancement, image segmentation, and feature extraction. Among them, medical image segmentation is the process of identifying and separating specific regions or structures in the image, providing more intuitive and accurate diagnostic basis for doctors. Medical images play an irreplaceable role in clinical diagnosis. Through medical images, doctors can intuitively observe the internal structure and pathological conditions of the human body, thereby making accurate diagnoses. In the fields of tumor detection and treatment, neuroimaging analysis, cardiovascular medicine, etc., medical image segmentation technology plays a crucial role. In addition, with the continuous progress and innovation of technology, medical images will play an even more important role in the future medical field.

[0004] Medical image analysis is a task highly dependent on doctors' experience and skills. However, even the most experienced doctors may make mistakes due to reasons such as fatigue and distraction. Moreover, in the era of big data, the quantity and complexity of medical image data are constantly increasing, and traditional medical image analysis methods have difficulty coping with this challenge. Summary of the Invention

[0005] In view of the above problems of the existing medical image analysis being a task highly dependent on doctors' experience and skills, however, even the most experienced doctors may make mistakes due to reasons such as fatigue and distraction, and in the era of big data, the quantity and complexity of medical image data are constantly increasing, and traditional medical image analysis methods have difficulty coping with this challenge, the present invention is proposed.

[0006] Therefore, the object of the present invention is to provide a key area detection system for medical images, which aims to: automatically identify and locate key anatomical structures or lesion areas in images, thereby greatly improving the diagnostic efficiency of doctors, providing important references for doctors' clinical decisions. By accurately identifying the lesion areas, doctors can more accurately evaluate the severity of the condition, formulate more effective treatment plans, and even predict the development trend of the disease, which is of great significance in improving the treatment effect and survival rate of patients. At the same time, through automated analysis, errors caused by human factors are reduced, and the reliability of diagnosis is improved.

[0007] To solve the above technical problems, the present invention provides the following technical solutions: A key area detection system for medical images includes an image acquisition module, which is used to ensure that the system can process various types of medical images and provide a basis for subsequent image processing and analysis.

[0008] An image preprocessing module, which is used to improve the readability of the image and the accuracy of subsequent processing.

[0009] A key area localization module, which realizes the core function of the system, that is, the automated detection of key areas.

[0010] A feature extraction module, which is used to provide basic data for subsequent classification, recognition, or analysis.

[0011] A classification and recognition module, which is used to classify and recognize the extracted features to determine the category or nature of the key area.

[0012] A result visualization and output module, which is used to display the results of detection and analysis in a visual form.

[0013] A database management and storage module, which is used to store and manage a large amount of medical image data and detection results.

[0014] A user interface module, which is used to provide a user interface for doctors to facilitate operation and setting.

[0015] As a preferred solution of the key area detection system for medical images of the present invention, wherein: the image acquisition module is mainly responsible for acquiring medical images from different data sources or sensors, and it includes an image acquisition device interface module and a data transmission and reception module.

[0016] As a preferred solution of the key area detection system for medical images of the present invention, wherein: the image preprocessing module can perform preliminary processing on the acquired medical images to improve the image quality and reduce noise, and it includes an image denoising module, an image enhancement module, and an image normalization module.

[0017] As a preferred solution of the key region detection system for medical images according to the present invention, wherein: the key region positioning module is responsible for positioning and identifying key regions of interest in medical images, and it includes a target detection algorithm module and a region segmentation technology module;

[0018] Target detection algorithm module;

[0019] In target detection, classifying the extracted features to determine the category of the object is an important step, which is usually achieved by a fully connected layer or a convolutional layer followed by a softmax function. The softmax function is used to convert the output into a probability distribution, and its calculation formula is:

[0020] softmax(z_i) = e(z_j)

[0021] where z_i is the i-th element of the input vector, and Σ_j e(z_j) is the sum. The output of the softmax function represents the probability of each category;

[0022] Calculation of gray mean and standard variance of the region segmentation technology module: For each pixel point (x, y) in the image, calculate the gray mean M(x, y) and standard variance d(x, y) within its neighborhood;

[0023] Adaptive threshold calculation: According to the gray mean and standard variance, calculate the adaptive threshold t(x, y) = α × d(x, y) + M(x, y), where α is a correction factor.

[0024] As a preferred solution of the key region detection system for medical images according to the present invention, wherein: the feature extraction module extracts and analyzes image features in the key region, and it includes a texture feature extraction module, a shape feature extraction module, and deep learning feature extraction;

[0025] Feature extraction module based on convolutional neural network. In a convolutional neural network (CNN), feature extraction is usually achieved through multiple convolutional layers, activation layers, and pooling layers. The convolutional layer performs local perception and feature extraction on the input image through a convolutional kernel (or filter);

[0026] The formula for convolution operation is: y(i, j) = ∑ m ∑ n x(i + m, j + n) · (m, n).

[0027] As a preferred solution of the key region detection system for medical images according to the present invention, wherein: the classification and recognition module classifies and recognizes the extracted features, and it includes a classifier training module and a classification and recognition algorithm module;

[0028] The classification and recognition module includes a Bayesian classifier, which is a classification method based on Bayes' theorem. It uses prior probability and conditional probability to calculate the posterior probability for classification.

[0029] Bayes' theorem:

[0030] Among them, P(A|B) is the probability that event A occurs under the condition that event B occurs (posterior probability), P(B|A) is the probability that event B occurs under the condition that event A occurs (conditional probability), P(A) is the prior probability that event A occurs, and P(B) is the probability that event B occurs.

[0031] For multi-classification problems, assuming there are n classes C1, C2... C n , for a given feature vector x, the Bayesian classifier selects the class with the maximum posterior probability as the prediction result.

[0032]

[0033] As a preferred solution of the key region detection system for medical images according to the present invention, wherein: the result visualization and output module displays the processing results of the detection system in a visual form, which includes a result visualization tool module and a report generator module.

[0034] As a preferred solution of the key region detection system for medical images according to the present invention, wherein: the database management and storage module is responsible for storing and managing medical image data and its processing results, which includes a database design module, a data storage module, and a data backup and recovery module.

[0035] As a preferred solution of the key region detection system for medical images according to the present invention, wherein: the user interface module provides a friendly operation interface and interaction method for users, which includes an interface design module, an interaction function implementation module, and a user permission management module.

[0036] To achieve the above object, the present invention provides the following technical solution: a detection method for a key region detection system of medical images, including the following steps:

[0037] S1. First, use a medical imaging device (such as CT, MRI, X-ray, etc.) to obtain the medical image of the patient, and ensure that the image quality meets the diagnostic requirements, including resolution, contrast, noise, etc.

[0038] S2. Then, perform denoising processing on the obtained medical image to reduce noise interference in the image, and then perform image enhancement to improve the contrast and clarity of the image for subsequent processing.

[0039] S3, perform image registration as needed to spatially align images at different times or of different modalities, and perform operations such as image cropping and scaling at the end to meet the requirements of subsequent processing flows;

[0040] S4, use image segmentation techniques such as threshold segmentation, region growing, level set methods, etc. to segment key regions (such as organs, tumors, etc.) in medical images;

[0041] S5, use methods such as shape analysis and feature matching to further localize and refine the segmented regions;

[0042] S6, extract features from the located key regions. These features may include shape features (such as area, perimeter, compactness, etc.), texture features (such as gray-level co-occurrence matrix, wavelet transform, etc.), and statistical features (such as mean, variance, histogram, etc.), and use deep learning techniques (such as convolutional neural network CNN) to automatically extract features. This method can learn more advanced and abstract feature representations;

[0043] S7, use machine learning algorithms (such as support vector machine SVM, random forest, neural network, etc.) to classify and identify the extracted features, and then determine the category of the key region (such as normal tissue, diseased tissue, etc.) according to the classification results;

[0044] S8, present the results of classification and identification to the user in a visual way, such as marking the positions and categories of key regions, generating a detailed detection report including information such as images of key regions, feature values, and classification results, and finally output the detection results to the user or doctor in electronic or paper form;

[0045] S9, finally establish and manage a medical image database to store information such as patients' medical images, feature data, and classification results, and provide functions such as data query, retrieval, and backup to ensure the security and integrity of the data. Finally, regularly maintain and update the database to meet new detection requirements and standards.

[0046] Advantages of the present invention: Through this system, key anatomical structures or diseased regions in images can be automatically identified and located, thus greatly improving the diagnostic efficiency of doctors, providing important references for doctors' clinical decisions. By accurately identifying diseased regions, doctors can more accurately assess the severity of the condition, formulate more effective treatment plans, and even predict the development trend of the disease, which is of great significance for improving the treatment effect and survival rate of patients. At the same time, through automated analysis, errors caused by human factors are reduced, and the reliability of diagnosis is improved. Description of the Drawings

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings. Among them:

[0048] Figure 1 It is a schematic diagram of the overall framework of the key area detection system for medical images of the present invention. Specific embodiments

[0049] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification.

[0050] In the following description, many specific details are set forth to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0051] Referring to Figure 1 , an embodiment of the present invention provides a key area detection system for medical images. This key area detection system for medical images includes an image acquisition module, which is responsible for acquiring image data from different medical imaging devices (such as CT, MRI, X-ray machines, etc.) to ensure that the system can process various types of medical images and provide a basis for subsequent image processing and analysis;

[0052] An image preprocessing module that performs preliminary processing on the acquired images, such as denoising, enhancing contrast, adjusting brightness, etc., to improve the image quality and is used to improve the readability of the images and the accuracy of subsequent processing;

[0053] A key area localization module that uses computer vision and machine learning algorithms to automatically identify and locate key anatomical structures or lesion areas in medical images, and adopts deep learning technologies such as object detection and semantic segmentation to achieve the core function of the system, that is, the automatic detection of key areas;

[0054] A feature extraction module that extracts useful feature information, such as shape, texture, color, etc., from the located key areas and is used to provide basic data for subsequent classification, recognition, or analysis;

[0055] A classification and recognition module that is used to classify and recognize the extracted features to determine the category or nature of the key areas, and adopts classification algorithms such as support vector machines, decision trees, and deep learning;

[0056] Result visualization and output module, which is used to display the results of detection and analysis in a visual form, such as drawing bounding boxes, labels, trajectories, etc., and provide output data or reports to facilitate doctors to quickly understand the detection results and make accurate diagnoses;

[0057] Database management and storage module, which is used to store and manage a large amount of medical image data and detection results, facilitating doctors to review historical cases for long-term tracking and research;

[0058] User interface module, which is used to provide a user interface to facilitate doctors to operate and set, improving the usability and user experience of the system.

[0059] The image acquisition module is mainly responsible for acquiring medical images from different data sources or sensors, and it includes an image acquisition device interface module, which is an interface for connecting and exchanging data with various medical imaging devices (such as CT, MRI, X-ray machines, etc.);

[0060] Data transmission and reception module, which is responsible for transmitting the acquired image data from the imaging device to the detection system and ensuring the integrity and accuracy of the data.

[0061] The image preprocessing module can perform preliminary processing on the acquired medical images to improve image quality and reduce noise. It includes an image denoising module, which removes noise and artifacts in the image to improve image clarity;

[0062] Image enhancement module, which enhances the details and readability of the image through techniques such as contrast stretching and sharpening filtering;

[0063] Image normalization module, which uniformly processes the size, resolution, gray value, etc. of the image for subsequent processing.

[0064] The key region localization module is responsible for locating and identifying key regions of interest in medical images. It includes an object detection algorithm module, a deep learning-based object detection network, which is used to quickly and accurately locate the target region in the image;

[0065] Object detection algorithm module;

[0066] In object detection, classifying the extracted features to determine the category of the object is an important step, which is usually achieved through a fully connected layer or a convolutional layer followed by a softmax function. The softmax function is used to convert the output into a probability distribution, and its calculation formula is:

[0067] softmax(z_i)=e(z_j)

[0068] Among them, z_i is the i-th element of the input vector, the sum of Σ_j e(z_j). The output of the softmax function represents the probability of each category;

[0069] Region segmentation technology module grayscale mean and standard deviation calculation: For each pixel point (x, y) in the image, calculate the grayscale mean M(x, y) and standard deviation d(x, y) within its neighborhood;

[0070] Adaptive threshold calculation: According to the grayscale mean and standard deviation, calculate the adaptive threshold t(x, y) = α × d(x, y) + M(x, y), where α is the correction factor;

[0071] The region segmentation technology module, such as threshold-based segmentation, region-based segmentation, edge-based segmentation, etc., is used to separate the target region from the background.

[0072] The feature extraction module extracts and analyzes the image features in the key region, which includes a texture feature extraction module that analyzes the texture information of the image and extracts useful texture features;

[0073] The feature extraction module based on a convolutional neural network. In a convolutional neural network (CNN), feature extraction is usually achieved through multiple convolutional layers, activation layers, and pooling layers. The convolutional layer performs local perception and feature extraction on the input image through a convolutional kernel (or filter);

[0074] The formula for the convolution operation is: y(i, j) = ∑ m ∑ n x(i + m, j + n) · (m, n)

[0075] The shape feature extraction module analyzes the shape of the target in the image and extracts shape features;

[0076] Deep learning feature extraction, using deep learning models such as convolutional neural networks to extract high-level features in the image.

[0077] The classification and recognition module classifies and recognizes the extracted features, which includes a classifier training module that uses image data of known categories to train the classifier so that it can accurately identify the target category;

[0078] The classification and recognition algorithm module, such as support vector machines, decision trees, K-nearest neighbor algorithms, etc., is used to classify and recognize the extracted features;

[0079] The classification and recognition module includes a Bayesian classifier, which is a classification method based on Bayes' theorem. It uses prior probabilities and conditional probabilities to calculate posterior probabilities for classification;

[0080] Bayes' theorem:

[0081] Among them, P(A|B) is the probability of event A occurring under the condition that event B occurs (posterior probability), P(B|A) is the probability of event B occurring under the condition that event A occurs (conditional probability), P(A) is the prior probability of event A occurring, and P(B) is the probability of event B occurring;

[0082] For multi-classification problems, assume there are n classes C1, C2... C n , for the given feature vector x, the Bayesian classifier selects the class with the maximum posterior probability as the prediction result

[0083]

[0084] The result visualization and output module displays the processing results of the detection system in a visual form, which includes a result visualization tool module, an image annotation tool, data visualization software, etc., for drawing target boxes, labels, trajectories, etc. on the image;

[0085] The report generator module generates a detailed report based on the processing results of the detection system, including information such as target location, category, attributes, etc.

[0086] The database management and storage module is responsible for storing and managing medical image data and its processing results, which includes a database design module to design a reasonable database structure for efficient storage and retrieval of image data;

[0087] The data storage module, relational databases, non-relational databases, distributed storage, etc., is used to store a large amount of image data and processing results;

[0088] The data backup and recovery module ensures the security and reliability of the database and prevents data loss or damage.

[0089] The user interface module provides a friendly operation interface and interaction method for users, which includes an interface design module to design an intuitive and easy-to-use operation interface for users to operate and view results;

[0090] The interaction function implementation module, functions such as image browsing, annotation, query, etc., to meet the actual needs of users

[0091] The user permission management module sets different permissions for different users to ensure the security of the system and the confidentiality of data.

[0092] A detection method for a key region detection system of medical images includes the following steps,

[0093] S1. First, use a medical imaging device (such as CT, MRI, X-ray, etc.) to obtain the patient's medical images, ensuring that the image quality meets the diagnostic requirements, including resolution, contrast, noise, etc.;

[0094] S2. Then, perform denoising processing on the obtained medical images to reduce noise interference in the images, and then perform image enhancement to improve the contrast and clarity of the images for subsequent processing;

[0095] S3. Perform image registration as needed to spatially align images taken at different times or of different modalities, and finally perform operations such as image cropping and scaling to meet the requirements of the subsequent processing flow;

[0096] S4. Use image segmentation techniques, such as threshold segmentation, region growing, level set method, etc., to segment key regions (such as organs, tumors, etc.) in the medical images;

[0097] S5. Use methods such as shape analysis and feature matching to further localize and refine the segmented regions;

[0098] S6. Extract features from the located key regions. These features may include shape features (such as area, perimeter, compactness, etc.), texture features (such as gray-level co-occurrence matrix, wavelet transform, etc.), and statistical features (such as mean, variance, histogram, etc.), and use deep learning techniques (such as convolutional neural network CNN) to automatically extract features. This method can learn more advanced and abstract feature representations;

[0099] S7. Use machine learning algorithms (such as support vector machine SVM, random forest, neural network, etc.) to classify and identify the extracted features, and then determine the category of the key regions (such as normal tissue, diseased tissue, etc.) according to the classification results;

[0100] S8. Present the results of classification and identification to the user in a visual way, such as marking the positions and categories of the key regions, generating a detailed detection report, including information such as images of the key regions, feature values, classification results, etc., and finally output the detection results to the user or doctor in electronic or paper form;

[0101] S9. Finally, establish and manage a medical image database for storing information such as the patient's medical images, feature data, classification results, etc., and provide functions such as data query, retrieval, backup, etc., to ensure the security and integrity of the data. Finally, perform regular maintenance and update on the database to adapt to new detection requirements and standards.

[0102] It should be noted that the above embodiments 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 preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A critical area detection system for medical images, characterized in that: include: Image acquisition module, used to ensure that the system can process various types of medical images and provide a basis for subsequent image processing and analysis; Image preprocessing module, used to improve the readability of images and the accuracy of subsequent processing; The key area positioning module realizes the core function of the system, namely the automatic detection of key areas; Feature extraction module, used to provide basic data for subsequent classification, recognition or analysis; A classification and recognition module is used to classify and recognize the extracted features to determine the category or nature of the key area; Result visualization and output module, used to display the detection and analysis results in a visual form; Database management and storage module, used to store and manage large amounts of medical imaging data and test results; The user interaction interface module is used to provide a user interaction interface to facilitate doctors to operate and set up.

2. The critical area detection system for medical images according to claim 1, characterized in that: The image acquisition module is mainly responsible for acquiring medical images from different data sources or sensors, and includes an image acquisition device interface module and a data transmission and receiving module.

3. The critical area detection system for medical images according to claim 2, characterized in that: The image preprocessing module can perform preliminary processing on the acquired medical image to improve image quality and reduce noise, and includes an image denoising module, an image enhancement module and an image standardization module.

4. The critical area detection system for medical images according to claim 3, characterized in that: The key region positioning module is responsible for locating and identifying key regions of interest in medical images, and includes a target detection algorithm module and a region segmentation technology module; Target detection algorithm module; In object detection, classifying the extracted features to determine the category of the object is an important step, which is usually achieved through a fully connected layer or a convolutional layer followed by a softmax function. The softmax function is used to convert the output into a probability distribution, and its calculation formula is: softmax(z_i)=e(z_j) Where z_i is the i-th element of the input vector, and the sum of Σ_j e(z_j). The output of the softmax function represents the probability of each category; Grayscale mean and standard deviation calculation of the region segmentation technology module: For each pixel (x, y) in the image, calculate the grayscale mean M(x, y) and standard deviation d(x, y) in its neighborhood; Adaptive threshold calculation: According to the grayscale mean and standard deviation, calculate the adaptive threshold t(x,y)=α×d(x,y)+M(x,y), where α is the correction factor.

5. The critical area detection system for medical images according to claim 4, characterized in that: The feature extraction module extracts and analyzes the image features in the key area, and includes a texture feature extraction module, a shape feature extraction module and a deep learning feature extraction module; Feature extraction module based on convolutional neural network. In convolutional neural network (CNN), feature extraction is usually achieved through multiple layers of convolutional layers, activation layers and pooling layers. The convolutional layer performs local perception and feature extraction on the input image through convolution kernels (or filters); The formula for convolution operation is: y(i,j)=∑ m ∑ n x(i+m,j+n)·(m,n).

6. The critical area detection system for medical images according to claim 5, characterized in that: The classification and recognition module classifies and recognizes the extracted features, and includes a classifier training module and a classification and recognition algorithm module; The classification and recognition module includes a Bayesian classifier, which is a classification method based on Bayes' theorem, which uses prior probability and conditional probability to calculate the posterior probability for classification; Bayes’ Theorem: Among them, P(A|B) is the probability of event A occurring under the condition that event B occurs (posterior probability), P(B|A) is the probability of event B occurring under the condition that event A occurs (conditional probability), P(A) is the prior probability of event A occurring, and P(B) is the probability of event B occurring; For multi-classification problems, suppose there are n categories C1, C2...C n , for a given feature vector x, the Bayesian classifier selects the category with the maximum posterior probability as the prediction result 7. The critical area detection system for medical images according to claim 6, characterized in that: The result visualization and output module displays the processing results of the detection system in a visual form, and includes a result visualization tool module and a report generator module.

8. The critical area detection system for medical images according to claim 7, characterized in that: The database management and storage module is responsible for storing and managing medical image data and processing results thereof, and includes a database design module, a data storage module, and a data backup and recovery module.

9. The critical area detection system for medical images according to claim 8, characterized in that: The user interaction interface module provides users with a friendly operation interface and interaction mode, and includes an interface design module, an interaction function implementation module and a user authority management module.

10. A detection method of a key area detection system for medical images, applied to the key area detection system for medical images as claimed in claim 9, characterized in that: The following steps are involved: S1, first use medical imaging equipment (such as CT, MRI, X-ray, etc.) to obtain medical images of patients and ensure that the image quality meets the diagnostic requirements, including resolution, contrast, noise, etc.; S2, then denoising the acquired medical image to reduce noise interference in the image, and then image enhancement to improve the contrast and clarity of the image for subsequent processing; S3, perform image registration as needed, spatially align images of different times or modalities, and finally perform image cropping, scaling and other operations to meet the requirements of subsequent processing procedures; S4, using image segmentation techniques, such as threshold segmentation, region growing, level set method, etc., to segment key areas (such as organs, tumors, etc.) in medical images; S5, using shape analysis, feature matching and other methods to further locate and refine the segmented area; S6, extract features from the located key areas. These features may include shape features (such as area, perimeter, compactness, etc.), texture features (such as gray-level co-occurrence matrix, wavelet transform, etc.) and statistical features (such as mean, variance, histogram, etc.), and use deep learning technology (such as convolutional neural network CNN) to automatically extract features. This method can learn more advanced and abstract feature representations; S7, using machine learning algorithms (such as support vector machine SVM, random forest, neural network, etc.) to classify and identify the extracted features, and then determine the category of the key area (such as normal tissue, diseased tissue, etc.) according to the classification results; S8, presenting the classification and recognition results to the user in a visual manner, such as marking the location and category of the key area, generating a detailed test report, including images of the key area, feature values, classification results and other information, and finally outputting the test results to the user or doctor in electronic or paper form; S9, finally, establish and manage the medical image database to store the patient's medical images, feature data, classification results and other information, and provide data query, retrieval, backup and other functions to ensure the security and integrity of the data. Finally, regularly maintain and update the database to adapt to new detection needs and standards.

Images