Breast magnetic resonance image processing and diagnosing method based on intelligent analysis

Abstract

The invention relates to the technical field of magnetic resonance image processing and diagnosis, in particular to a breast magnetic resonance image processing and diagnosis method based on intelligent analysis, which comprises the following steps of: 1, establishing a personal information file: establishing a user personal health information file through terminal equipment, establishing a user individual mammary gland magnetic resonance image database associated with the mammary gland magnetic resonance shared data center; 2, acquiring a mammary gland magnetic resonance image: acquiring the mammary gland magnetic resonance image through mammary gland magnetic resonance equipment, and storing the mammary gland magnetic resonance image to a user personal mammary gland magnetic resonance image database; according to the method, the processing diagnosis of establishing the linear relation between the breast magnetic resonance image random variable and the characteristic value is objective and simple; dynamic enhanced scanning of the mammary gland magnetic resonance image is achieved, the image is clearer and more accurate, analysis, processing diagnosis and prediction analysis of the mammary gland magnetic resonance image of the user are diagramed, and the personal mammary gland health of the user can be dynamically tracked.

Description

Technical Field

[0001] The invention relates to the technical field of magnetic resonance image processing and diagnosis, and in particular to a method for processing and diagnosing breast magnetic resonance images based on intelligent analysis. Background Art

[0002] The incidence of breast cancer has been at the forefront of all types of malignant tumors for many years. The age of onset is mostly 40-60 years old, menopausal and perimenopausal women, and there has been a trend of younger age in recent years. Early detection, early diagnosis and early treatment of breast cancer play a decisive role in improving the survival rate and quality of life of patients. Breast MRI is a new medical imaging technology that uses the principle of nuclear magnetic resonance to act on breast tissue through a strong magnetic field to generate magnetic resonance signals and collect them multiple times to form images for diagnosis. Its imaging is clearer and the diagnosis is more accurate. It can display the lesions intuitively and stereoscopically from all directions and angles. It has high resolution for small lesions and is mainly used to display the overall state of the breast and distinguish between benign and malignant breast occupying masses. It has a high sensitivity for judging whether there is obvious metastasis of axillary lymph nodes and invasive breast cancer. For people suspected of breast cancer, it is generally a must-do examination.

[0003] At present, breast magnetic resonance images are processed and diagnosed by professional doctors, which has a large base and a heavy workload, leading to a shortage of medical resources and an increase in the medical burden. At the same time, it causes inconvenience to breast magnetic resonance users in seeking medical treatment. In addition, breast magnetic resonance image data is single and collaboration is difficult. In order to solve the above-mentioned problems, the present invention proposes a processing and diagnosis method for breast magnetic resonance images based on intelligent analysis. Summary of the invention

[0004] In view of the problems in the prior art, the present invention provides a method for processing and diagnosing breast magnetic resonance images based on intelligent analysis.

[0005] The technical solution adopted by the present invention to solve the technical problem is: the processing and diagnosis method comprises the following steps: Step 1: Establish a personal information file: Establish a user's personal health information file through a terminal device, and establish a user's personal breast MRI image database associated with a breast MRI shared data center; Step 2, obtaining a breast magnetic resonance image: obtaining a breast magnetic resonance image through a breast magnetic resonance device, and storing the breast magnetic resonance image in a user's personal breast magnetic resonance image database; Step 3, model diagnosis: Based on the breast magnetic resonance image processing diagnosis model, the acquired breast magnetic resonance image is imported into the breast magnetic resonance image processing diagnosis model, and the breast magnetic resonance image intelligent analysis and processing diagnosis is performed to obtain breast magnetic resonance image analysis and processing diagnosis data; Step 4, prediction analysis: based on the analysis and processing diagnosis results of the breast MRI image processing diagnosis model, the breast MRI image lesions are initially identified and stored in the user's personal archive database, and the breast MRI images with the initial identification are fed back to the breast MRI image prediction analysis model, and the benign and malignant nature of the breast lesions in the breast MRI images are predicted and analyzed by the breast MRI image prediction analysis model to obtain breast MRI image prediction analysis data; Step 5. Personal log: Graph the user's breast MRI image analysis, processing diagnosis and predictive analysis, and establish the user's personal breast MRI image dynamic log for dynamic tracking of the user's personal breast health.

[0006] Specifically, the construction of the breast magnetic resonance image processing diagnosis model is based on big data technology and a breast magnetic resonance shared data center, and the target set data and test data for breast magnetic resonance image processing diagnosis are obtained. The target set data and the test data are preprocessed, and null value data for the target set data and the test data are proposed to obtain breast magnetic resonance image processing diagnosis target feature data and test set data. Based on the improved stepwise regression and convolutional neural network algorithms of deep learning, the breast magnetic resonance image processing diagnosis target feature data is input into the neural network model to obtain the breast magnetic resonance image processing diagnosis target feature data, and the test set data is input into the breast magnetic resonance image processing diagnosis target feature data to analyze the sensitivity and specificity of the model.

[0007] Specifically, the breast magnetic resonance image processing and diagnosis model utilizes the technology of neural network model for machine learning and prediction. By constructing a multi-level neural network structure, it automatically extracts useful feature information from a large amount of data in the breast magnetic resonance shared data center, thereby realizing automated processing of classification, regression or other complex tasks of the target set data. It can automatically extract feature information of the breast from breast magnetic resonance images, thereby realizing accurate recognition and reading of breast magnetic resonance images.

[0008] Specifically, the breast MRI image acquisition can be carried out through an online platform for breast MRI appointment. The user registers and logs in to the online platform, fills in the user's personal information and appointment time, and the online platform automatically matches the user with a specific breast MRI location and time based on the user's appointment, sends the user an appointment confirmation message and a reminder of preparation precautions. The user undergoes a breast MRI examination as scheduled, obtains breast MRI images, and stores them in a database.

[0009] Specifically, the breast magnetic resonance image adopts conventional dynamic enhancement scanning to DCE dynamic enhancement quantitative perfusion scanning mode. After setting the scanning sequence, the corresponding data type, post-processing parameters, alignment sequence and calculation scheme, post-processing model and parameters are set in this setting menu. One phase sequence of dynamic enhancement is automatically selected as a reference, and other phase sequences are motion corrected. The pre-enhancement morphological sequence and the dynamic enhancement sequence are aligned. According to the predefined analysis model and the predefined volume, the corresponding parameter map is analyzed and calculated. Finally, the calculation is completed according to the predefined workflow, and the result reading step is directly entered to outline the area of ​​interest to obtain the result.

[0010] Specifically, the processing and diagnosis method also includes a well-trained radiologist or technician performing a personnel diagnosis and analysis based on the acquired breast magnetic resonance images.

[0011] Specifically, the model diagnosis also includes constructing a neural network model and using digital pathology images to accurately predict the classification of triple-negative breast cancer, and through a pathological HE-stained section, clarify the molecular classification and key treatment targets of triple-negative breast cancer within a few hours.

[0012] Specifically, the user's personal breast magnetic resonance image database is provided with an encryption module, which can encrypt and hide the user's personal data, so that the access to the user's personal breast magnetic resonance image database is protected by encryption authority.

[0013] Beneficial effects of the present invention: (1) The present invention discloses a method for processing and diagnosing breast MRI images based on intelligent analysis. The present invention adopts a breast MRI image processing and diagnosis model to perform intelligent analysis and processing and diagnosis of breast MRI images, and predicts and analyzes the benign and malignant nature of breast lesions in breast MRI images through a breast MRI image prediction and analysis model. The breast MRI image processing and diagnosis model is constructed by acquiring breast MRI image data through big data technology, so that the breast MRI image processing and diagnosis model is established, and the processing and diagnosis of the linear relationship between the random variable and the eigenvalue of the breast MRI image is established, which is objective and simple; the breast MRI image is dynamically enhanced and scanned, and the image is clearer and more accurate; (2) The present invention discloses a method for processing and diagnosing breast MRI images based on intelligent analysis. The present invention establishes a user's personal health information file, charts the obtained user's breast MRI image analysis, processing diagnosis and predictive analysis, and establishes a user's personal breast MRI image dynamic log, which can track the user's personal breast health status. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0015] Figure 1 The present invention provides a flow chart of a method for processing and diagnosing breast magnetic resonance images based on intelligent analysis. DETAILED DESCRIPTION

[0016] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0017] like Figure 1 As shown, a processing and diagnosis method of breast magnetic resonance images based on intelligent analysis of the present invention comprises the following steps: Step 1: Establish a personal information file: Establish a user's personal health information file through a terminal device, and establish a user's personal breast MRI image database associated with a breast MRI shared data center; Step 2, obtaining a breast magnetic resonance image: obtaining a breast magnetic resonance image through a breast magnetic resonance device, and storing the breast magnetic resonance image in a user's personal breast magnetic resonance image database; Step 3, model diagnosis: Based on the breast magnetic resonance image processing diagnosis model, the acquired breast magnetic resonance image is imported into the breast magnetic resonance image processing diagnosis model, and the breast magnetic resonance image intelligent analysis and processing diagnosis is performed to obtain breast magnetic resonance image analysis and processing diagnosis data; Step 4, prediction analysis: based on the analysis and processing diagnosis results of the breast MRI image processing diagnosis model, the breast MRI image lesions are initially identified and stored in the user's personal archive database, and the breast MRI images with the initial identification are fed back to the breast MRI image prediction analysis model, and the benign and malignant nature of the breast lesions in the breast MRI images are predicted and analyzed by the breast MRI image prediction analysis model to obtain breast MRI image prediction analysis data; Step 5. Personal log: Graph the user's breast MRI image analysis, processing diagnosis and predictive analysis, and establish the user's personal breast MRI image dynamic log for dynamic tracking of the user's personal breast health.

[0018] Specifically, the construction of the breast magnetic resonance image processing diagnosis model is based on big data technology and a breast magnetic resonance shared data center, and the target set data and test data for breast magnetic resonance image processing diagnosis are obtained. The target set data and the test data are preprocessed, and null value data for the target set data and the test data are proposed to obtain breast magnetic resonance image processing diagnosis target feature data and test set data. Based on the improved stepwise regression and convolutional neural network algorithms of deep learning, the breast magnetic resonance image processing diagnosis target feature data is input into the neural network model to obtain the breast magnetic resonance image processing diagnosis target feature data, and the test set data is input into the breast magnetic resonance image processing diagnosis target feature data. The sensitivity and specificity of the model are analyzed, and the breast magnetic resonance image processing diagnosis model is accurate.

[0019] Specifically, the breast magnetic resonance image processing and diagnosis model utilizes the technology of neural network model for machine learning and prediction. By constructing a multi-level neural network structure, it automatically extracts useful feature information from a large amount of data in the breast magnetic resonance shared data center, thereby realizing automated processing of classification, regression or other complex tasks of the target set data. It can automatically extract feature information of the breast from breast magnetic resonance images, thereby realizing accurate recognition and reading of breast magnetic resonance images.

[0020] Specifically, the breast MRI image acquisition can be carried out through an online platform for breast MRI appointment. The user registers and logs in to the online platform, fills in the user's personal information and appointment time, and the online platform automatically matches the user with a specific breast MRI location and time based on the user's appointment, sends the user an appointment confirmation message and a reminder of preparation precautions. The user undergoes a breast MRI examination as scheduled, obtains breast MRI images, and stores them in a database.

[0021] Specifically, the breast magnetic resonance image adopts conventional dynamic enhancement tracing to DCE dynamic enhancement quantitative perfusion scanning mode, and the target organ is analyzed from qualitative to semi-quantitative and then to quantitative results through Tissue 4D post-processing. The parameters of AT\TTP\WI\WO\AUC can be used to semi-quantitatively judge the degree of enhancement and the enhancement process. The quantitative parameters of ktrans\Ve\kep\iAUC can be calculated by converting the signal intensity into the concentration of the contrast agent and then fitting it with the corresponding pharmacokinetic model; the acquisition of quantitative parameters mainly depends on the selection of a drug metabolism model that matches the tissue well during the examination process. The appropriate model is selected to match the blood flow of the lesion or tissue as much as possible, the uptake and distribution of the contrast agent in the lesion, and the signal emphasis time curve that matches the blood flow of the area of ​​interest is obtained, so as to obtain more accurate parameters. The pharmacokinetic model used is the Tofts two-chamber model, one chamber is the intravascular space, and the other chamber is the extravascular extracellular space. When the intravascular space and the extravascular extracellular space reach - It is a dynamic equilibrium process. Substances will penetrate from the blood vessels into the extracellular space outside the blood vessels, and substances will also return from the extracellular space outside the blood vessels to the blood vessels, thus reflecting the vascular permeability. The system will match the AIF arterial input function. During post-processing, you need to select an AIF, and the system will automatically apply it to calculate the Ktrans and Kep values. After setting the scanning sequence, set the corresponding data type, post-processing parameters, alignment sequence and calculation scheme, post-processing model and parameters in this setting menu. Automatically select one of the dynamic enhancement sequences as a reference, and the other sequences and motion correction, pre-enhancement morphological sequence and dynamic enhancement sequence for registration. According to the predefined analysis model and predefined volume, analyze and calculate the corresponding parameter map. Finally, the calculation ends according to the predefined workflow, and directly enters the result reading step to outline the area of ​​interest to get the result.

[0022] Specifically, the processing and diagnosis method also includes a well-trained radiologist or technician performing personnel diagnosis and analysis based on the acquired breast magnetic resonance images to obtain personnel diagnosis and analysis data, and inputting the personnel diagnosis and analysis data and the model diagnosis and analysis data into the breast magnetic resonance image diagnosis and analysis system for comparison.

[0023] Specifically, the model diagnosis also includes constructing a neural network model and using digital pathology images to accurately predict the classification of triple-negative breast cancer, and through a pathological HE-stained section, clarify the molecular classification and key treatment targets of triple-negative breast cancer within a few hours.

[0024] Specifically, the user's personal breast magnetic resonance image database is provided with an encryption module, which can encrypt and hide the user's personal data, so that the access to the user's personal breast magnetic resonance image database has encryption authority protection, thereby improving the security of the user's personal information.

[0025] During use, the user establishes a user's personal health information file through the terminal device, establishes a user's personal breast MRI image database associated with the breast MRI shared data center, the user makes a breast MRI appointment through the online platform, fills in the user's personal information and appointment time, the online platform automatically matches the user with a specific breast MRI location and time based on the user's appointment, sends the user an appointment confirmation message and a reminder of preparation precautions, the user undergoes a breast MRI examination as scheduled, obtains breast MRI images through the breast MRI equipment, the breast MRI images are stored in the user's personal breast MRI image database, the obtained breast MRI images are imported into the breast MRI image processing and diagnosis model, the breast MRI images are intelligently analyzed and processed and diagnosed, and breast MRI image analysis and processing diagnosis data are obtained, based on the analysis and processing diagnosis results of the breast MRI image processing and diagnosis model, the breast MRI image lesions are initially identified and stored. To the user's personal archive database, the initially marked breast MRI image is fed back to the breast MRI image prediction and analysis model, the breast MRI image prediction and analysis model is used to predict and analyze the benign and malignant nature of breast lesions in the breast MRI image, and breast MRI image prediction and analysis data are obtained. A well-trained radiologist or technician performs personnel diagnosis and analysis based on the acquired breast MRI image to obtain personnel diagnosis and analysis data, and the personnel diagnosis and analysis data are input into the breast MRI image diagnosis and analysis system for comparison, and the user's breast MRI image analysis, processing diagnosis and prediction analysis are charted, and a user's personal breast MRI image dynamic log is established for dynamic tracking of the user's personal breast health, so that the breast MRI image processing and diagnosis are modeled, and a processing and diagnosis of the linear relationship between the random variable and the eigenvalue of the breast MRI image is established, which is objective and simple; dynamic enhanced scanning of breast MRI images makes the image clearer and more accurate.

[0026] The above shows and describes 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 above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for processing and diagnosing breast magnetic resonance images based on intelligent analysis, characterized in that: The treatment and diagnosis method comprises the following steps: Step 1: Establish a personal information file: Establish a user's personal health information file through a terminal device, and establish a user's personal breast MRI image database associated with a breast MRI shared data center; Step 2, obtaining a breast magnetic resonance image: obtaining a breast magnetic resonance image through a breast magnetic resonance device, and storing the breast magnetic resonance image in a user's personal breast magnetic resonance image database; Step 3, model diagnosis: Based on the breast magnetic resonance image processing diagnosis model, the acquired breast magnetic resonance image is imported into the breast magnetic resonance image processing diagnosis model, and the breast magnetic resonance image intelligent analysis and processing diagnosis is performed to obtain breast magnetic resonance image analysis and processing diagnosis data; Step 4, prediction analysis: based on the analysis and processing diagnosis results of the breast MRI image processing diagnosis model, the breast MRI image lesions are initially identified and stored in the user's personal archive database, and the breast MRI images with the initial identification are fed back to the breast MRI image prediction analysis model, and the benign and malignant nature of the breast lesions in the breast MRI images are predicted and analyzed by the breast MRI image prediction analysis model to obtain breast MRI image prediction analysis data; Step 5. Personal log: Graph the user's breast MRI image analysis, processing diagnosis and predictive analysis, and establish the user's personal breast MRI image dynamic log for dynamic tracking of the user's personal breast health.

2. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 1, characterized in that: The construction of the breast magnetic resonance image processing diagnosis model is based on big data technology and a breast magnetic resonance shared data center. Target set data and test data for breast magnetic resonance image processing diagnosis are obtained, the target set data and the test data are preprocessed, and null value data for the target set data and the test data are proposed to obtain target feature data and test set data for breast magnetic resonance image processing diagnosis. Based on an improved stepwise regression and convolutional neural network algorithm of deep learning, the target feature data for breast magnetic resonance image processing diagnosis is input into a neural network model to obtain target feature data for breast magnetic resonance image processing diagnosis. The test set data is input into the target feature data for breast magnetic resonance image processing diagnosis, and the sensitivity and specificity of the model are analyzed.

3. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 2, characterized in that: The breast magnetic resonance image processing and diagnosis model utilizes the technology of neural network model for machine learning and prediction. By constructing a multi-level neural network structure, useful feature information is automatically extracted from a large amount of data in the breast magnetic resonance shared data center, thereby realizing automatic processing of classification, regression or other complex tasks of the target set data. The feature information of the breast can be automatically extracted from the breast magnetic resonance image, thereby realizing accurate recognition and reading of the breast magnetic resonance image.

4. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 1, characterized in that: The breast MRI image acquisition can be carried out through an online platform for breast MRI reservation. The user registers and logs in to the online platform, fills in the user's personal information and reservation time, and the online platform automatically matches the user with a specific breast MRI location and time based on the user's reservation, sends the user reservation confirmation information and preparation precautions reminder, and the user performs a breast MRI examination according to the reservation, obtains breast MRI images, and stores them in a database.

5. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 1, characterized in that: The breast magnetic resonance image adopts conventional dynamic enhancement scanning to DCE dynamic enhancement quantitative perfusion scanning mode. After setting the scanning sequence, the corresponding data type, post-processing parameters, alignment sequence and calculation scheme, post-processing model and parameters are set in this setting menu. One phase sequence of dynamic enhancement is automatically selected as a reference, and other phase sequences are aligned with motion correction, pre-enhancement morphological sequence and dynamic enhancement sequence. According to the predefined analysis model and predefined volume, the corresponding parameter map is analyzed and calculated. Finally, the calculation is completed according to the predefined workflow, and the result reading step is directly entered to outline the area of ​​interest to obtain the result.

6. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 1, characterized in that: The processing and diagnosis method also includes a well-trained radiologist or technician performing a personnel diagnosis analysis based on the acquired breast magnetic resonance images.

7. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 1, characterized in that: The model diagnosis also includes building a neural network model and using digital pathology images to accurately predict the classification of triple-negative breast cancer, and through a pathology HE-stained section, clarify the molecular classification and key treatment targets of triple-negative breast cancer within a few hours.

8. The method for processing and diagnosing breast magnetic resonance images based on intelligent analysis according to claim 1, characterized in that: The user's personal breast magnetic resonance image database is provided with an encryption module, which can encrypt and hide the user's personal data, so that the access to the user's personal breast magnetic resonance image database has encryption authority protection.

Images