Medical image analysis method based on multi-column neural network

Abstract

The invention relates to the field of medical image analysis, in particular to a medical image analysis method based on a multi-column neural network, and the method comprises the steps: carrying out the data collection through an image collection module, transmitting the collected data to a control module, processing the image information through the control module, and carrying out the display processing on a display device. The feature information in the image is extracted through the extraction module, the position of the lesion tissue is positioned, the position of the lesion tissue in the image is amplified, meanwhile, definition restoration is conducted on the amplified image through computer software, and the definition of the amplified image is improved; the beneficial effects of the invention are that the method extracts the collected image information, locates the position of the lesion tissue in the image, amplifies the position, and corrects the amplified image, thereby improving the definition of the amplified image, facilitating a worker to clear and watch the lesion tissue in the image, and improving the work efficiency. And a worker is assisted to quickly diagnose a lesion reason.

Description

Technical Field

[0001] This invention relates to the field of medical image analysis, specifically a medical image analysis method based on a multi-column neural network. Background Technology

[0002] Computer-aided analysis of medical images utilizes advanced computer hardware and software systems to analyze and process digital radiographic images to identify and detect lesion features. The results serve as a "second opinion" for the diagnostic physician's reference. Currently, when using computer hardware and software systems to analyze images, the image information is mostly displayed on display devices for staff to view. However, due to the large size of the acquired images, the clarity is significantly reduced when staff view specific lesions, which is not conducive to viewing. Summary of the Invention

[0003] To address the problems in the existing technology, this invention provides a medical image analysis method based on a multi-column neural network.

[0004] The technical solution adopted by this invention to solve its technical problem is: a medical image analysis method based on a multi-column neural network, comprising:

[0005] Data is acquired through an image acquisition module and sent to a control module. The control module processes the image information and displays the processed data on a display device. An extraction module extracts feature information from the image, locates the diseased tissue, and magnifies the area of ​​the diseased tissue in the image. Simultaneously, computer software restores the clarity of the magnified image. Based on the extracted feature information, relevant medical records are searched in a database. Based on image similarity, the medical record with the closest image information is selected from numerous records, and the symptoms of the diseased tissue are analyzed. The diagnostic data is displayed below the image information for staff to view. Once the staff confirms the symptoms, the image information is numbered, stored in the data storage module, and simultaneously uploaded to a cloud server.

[0006] Specifically, when acquiring images, one can manually take pictures using a camera and then upload them to a computer.

[0007] Specifically, multiple display devices can be used to display images and data, and image information and data information can be displayed separately for staff reference.

[0008] Specifically, when displayed on a display device, the 3D model building module can construct a 3D model of the diseased tissue, making it easier for staff to understand the condition of the diseased tissue and the surrounding tissue.

[0009] Specifically, the collected image information can be numbered before being stored in the storage module and uploaded to the cloud server to avoid duplicate patient name information, which would affect subsequent retrieval and understanding.

[0010] Specifically, after extracting feature information from the image, the control module can perform preliminary screening in the database based on the feature information, and then select the medical record with the highest similarity to the extracted image from the screened medical record images.

[0011] Specifically, the image information is encrypted using an encryption module during storage to prevent information leakage.

[0012] The beneficial effects of this invention are as follows: This invention extracts the information from the acquired image, locates the position of the lesion in the image, then magnifies this position and corrects the magnified image to improve the clarity of the magnified image, making it easier for staff to view the lesion and assisting them in quickly diagnosing the cause of the lesion. Attached Figure Description

[0013] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0014] Figure 1 A flowchart provided for this invention. Detailed Implementation

[0015] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0016] like Figure 1 As shown, the medical image analysis method based on a multi-column neural network according to the present invention includes:

[0017] Data is acquired through an image acquisition module and sent to a control module. The control module processes the image information and displays the processed data on a display device. An extraction module extracts feature information from the image, locates the diseased tissue, and magnifies the area of ​​the diseased tissue in the image. Simultaneously, computer software restores the clarity of the magnified image. Based on the extracted feature information, relevant medical records are searched in a database. Based on image similarity, the medical record with the closest image information is selected from numerous records, and the symptoms of the diseased tissue are analyzed. The diagnostic data is displayed below the image information for staff to view. Once the staff confirms the symptoms, the image information is numbered, stored in the data storage module, and simultaneously uploaded to a cloud server.

[0018] Specifically, during image acquisition, images can be taken manually using photographic tools and then uploaded to a computer. Multiple display devices can be used to display images and data, allowing for separate display of image and data information for staff reference. When displaying images on these devices, a 3D modeling module can construct a 3D model of the lesion tissue, facilitating staff understanding of the lesion and surrounding tissue. Acquired images can be numbered before being stored in a storage module and uploaded to a cloud server, preventing duplicate patient names that could hinder later retrieval. After extracting feature information from the images, the control module can perform initial screening in a database based on these features, then select the medical records with the highest similarity to the extracted images from the remaining images. Finally, image information is encrypted during storage using an encryption module to prevent information leakage.

[0019] In use, data is acquired through the image acquisition module and sent to the control module. The control module processes the image information and displays it on the display device. The extraction module extracts feature information from the image to locate the lesion, magnifies the location of the lesion in the image, and uses computer software to restore the clarity of the magnified image. Based on the extracted feature information, relevant medical records are searched in the database. Based on image similarity, the medical record with the closest image information is found among many medical records. The symptoms of the lesion are analyzed, and the diagnostic data is displayed below the image information for staff to view. Once the staff confirms the symptoms, the image information is numbered, stored in the data storage module, and uploaded to the cloud server.

[0020] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of protection claimed by the present invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A medical image analysis method based on a multi-column neural network, comprising: Data is acquired through the image acquisition module, and the acquired data is sent to the control module. The control module processes the image information and displays it on the display device. The feature information in the image is extracted by the extraction module to locate the location of the lesion tissue. The location of the lesion tissue in the image is magnified, and the clarity of the magnified image is restored by computer software to improve the clarity of the magnified image. Based on the extracted feature information, relevant medical records are searched in the database. Based on image similarity, the medical record with the closest image information is found among many medical records, and the symptoms of the lesion tissue are analyzed. The diagnostic data is displayed below the image information for staff to view. Once the staff confirms the symptoms, the image information is numbered, stored in the data storage module, and simultaneously uploaded to the cloud server.

2. The medical image analysis method based on a multi-column neural network according to claim 1, characterized in that: When acquiring images, you can manually take pictures using a camera and then upload them to a computer.

3. The medical image analysis method based on a multi-column neural network according to claim 1, characterized in that: Multiple display devices can be used to display images and data, and image information and data information can be displayed separately for staff reference.

4. The medical image analysis method based on a multi-column neural network according to claim 1, characterized in that: When displayed on a display device, the 3D model building module can construct a 3D model of the lesion tissue, making it easier for staff to understand the lesion tissue and the surrounding tissue.

5. The medical image analysis method based on a multi-column neural network according to claim 1, characterized in that: The collected image information can be numbered before being stored in the storage module and uploaded to the cloud server to avoid duplicate patient name information, which would affect subsequent retrieval and understanding.

6. The medical image analysis method based on a multi-column neural network according to claim 1, characterized in that: After extracting feature information from the image, the control module can perform preliminary screening in the database based on the feature information, and then select the medical record with the highest similarity to the extracted image from the screened medical record images.

7. The medical image analysis method based on a multi-column neural network according to claim 1, characterized in that: The image information is encrypted using an encryption module during storage to prevent information leakage.

Images