Intelligent analysis and diagnosis auxiliary system for medical image data

Abstract

The invention discloses an intelligent analysis and diagnosis auxiliary system for medical image data, and belongs to the technical field of medical image processing and analysis, and the system comprises an image data obtaining module which is used for obtaining head CT image data of a preset patient, and marking the head CT image data as initial head CT image data; analyzing a lesion form, a lesion position and a lesion density of the brain of the patient based on the head CT image data, and evaluating lesion properties; the image data fusion processing unit responds to the evaluation result of the lesion property and is used for acquiring edge image features of the lesion form according to the image data; the image data fusion processing unit comprises a feature acquisition module, an analysis module and a comparison module. Through automatic analysis, dynamic monitoring, accurate evaluation, prediction and early warning and monitoring strategy optimization, the diagnosis efficiency and accuracy are improved, comprehensive diagnosis support is provided for doctors, and the medical service quality can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing and analysis, and particularly to an intelligent analysis and diagnostic assistance system for medical image data. Background Art

[0002] In the modern medical system, medical images play an important role in disease diagnosis. Imaging examinations such as X-ray, CT, MRI, and ultrasound can provide doctors with rich lesion information. Medical imaging examinations have become an indispensable key link in disease diagnosis and treatment evaluation. With the rapid development of medical technology and the increasing popularity of imaging equipment, imaging data has grown explosively. However, the quality control of imaging data faces challenges. For example, the traditional method of manually sampling and auditing medical images for quality control is inefficient, difficult to control a large amount of imaging data, and subjective, prone to missing potential quality hazards. To overcome these challenges and meet the clinical demand for accurate imaging diagnosis, an AI intelligent assistance system for medical imaging data has emerged, aiming to improve the quality of medical services with the power of artificial intelligence.

[0003] In the existing research on AI intelligent assistance systems for medical imaging data, the application document with the application number CN202411647493.8 provides a medical image assisted analysis method and system based on artificial intelligence. The technical solution includes: obtaining a specified number of medical images of the target part; selecting a reference medical image from the prior medical images; aligning the specified number of medical images in a unified coordinate system and selecting a reference boundary; performing segmentation, calculating the similarity between two sub-segmentations for the segmentation at the same position; determining a possible lesion area according to the trend of the calculated similarity; determining the shadow range of the possible lesion area, and extracting the shadow features of multiple medical images in chronological order. This technical solution combines image processing and artificial intelligence technologies to achieve the analysis of local shadow areas of medical images, improving the diagnostic efficiency of medical images to a certain extent.

[0004] However, for head CT images of patients, the morphology, position, and density of brain lesions vary greatly due to individual differences, making it difficult for the AI model to accurately identify and classify, thereby reducing the analysis accuracy of the AI model and making it possible that some early or subtle brain lesions (such as small cerebral hemorrhages and early cerebral infarctions) may be difficult to detect by the existing AI models. Summary of the Invention

[0005] In view of the above problems existing in the current technical field of medical image processing and analysis, the present invention is proposed.

[0006] Therefore, one of the objectives of the present invention is to provide an intelligent analysis and diagnostic assistance system for medical imaging data, which improves the diagnostic efficiency and accuracy through automated analysis, dynamic monitoring, precise evaluation, prediction and early warning, and optimization of monitoring strategies, provides comprehensive diagnostic support for doctors, helps improve the quality of medical services, and helps doctors formulate treatment plans in advance.

[0007] To solve the above technical problems, the present invention provides the following technical solutions:

[0008] The present invention provides an intelligent analysis and diagnostic assistance system for medical imaging data, including:

[0009] An image data acquisition module, configured to acquire head CT image data of a preset patient, and mark the head CT image data as initial head CT image data; analyze the lesion morphology, lesion location, and lesion density of the patient's brain based on the head CT image data, and evaluate the lesion nature;

[0010] An image data fusion processing unit, which responds to the evaluation result of the lesion nature, and is configured to obtain the edge image features of the lesion morphology according to the image data; the image data fusion processing unit includes a feature acquisition module, an analysis module, and a comparison module;

[0011] The feature acquisition module is configured to acquire the edge image features of the lesion morphology, and its acquisition method includes decomposing and reconstructing the edge image of the lesion morphology based on the wavelet transform method to extract the edge image features;

[0012] The analysis module responds to the extracted edge image features, and is configured to set a gray threshold according to the edge image features, and distinguish the pixels in the edge image into edge pixels and non-edge pixels; mark the non-edge pixels as reference edge pixels; and update the evaluation result of the lesion nature according to the change of the reference edge pixels.

[0013] The comparison module marks the head CT image data obtained for the preset patient in the future period as reference head CT image data; analyzes the change characteristics of the reference edge pixels based on the edge pixels of the reference head CT image data. If the change characteristics of the reference edge pixels are the same as those of the edge pixels, the system determines that the area corresponding to the lesion location of the preset patient shows a trend of diffusion change, and at the same time updates the evaluation result of the lesion nature to deterioration; otherwise, it does not make a determination.

[0014] As a preferred embodiment of the present invention, wherein: the image data fusion processing unit further includes a calculation module, and the calculation module is configured to calculate the time difference between the two acquisitions of the head CT image data based on the reference head CT image data and the initial head CT image data, and calculate the daily diffusion rate of the lesion location of the preset patient according to the time difference.

[0015] As a preferred embodiment of the present invention, wherein: according to the calculated diffusion rate, the reference edge pixels are divided into edge points in the following way; wherein, represents the nth edge point, and within the time difference, the diffusion characteristics of different edge points are analyzed, an edge point with the largest diffusion area is obtained, and the symptom manifestations and / or disease progression of the preset patient are analyzed based on the edge point.

[0016] As a preferred embodiment of the present invention, wherein: the edge point with the largest diffusion area is marked as the critical edge point, the diffusion direction of the critical edge point is analyzed, and the diffusion rate of each other edge point is calculated according to the diffusion characteristics of the critical edge point within the time difference, and is calculated according to the following formula:

[0017] wherein, p j represents the pth diffusion rate calculated for each edge point at the jth time point within the time difference;

[0018] In the formula, u represents the gray value of each edge point when the calculated pth diffusion rate is reached, t represents time, and the time is expressed as a percentage, and this time percentage represents the percentage of the time point when the calculated pth diffusion rate is reached in the time difference; d is the diffusion coefficient, representing the diffusion rate; v 2 represents the second-order derivative in space.

[0019] As a preferred embodiment of the present invention, wherein: according to the calculation result, the number of other edge points except the critical edge point is counted, and among the number, 5 to 10 edge points with the slowest diffusion rate are counted according to the calculation result, and the proportion of the edge points in the number is calculated. If the proportion of the edge points exceeds half of the number, the system determines that the lesion speed of the preset patient is slow; otherwise, it is not determined.

[0020] As a preferred embodiment of the present invention, if the system determines that the lesion growth rate of the preset patient is slow, a monitoring period at least twice the time difference is preset. Starting from the first day of the preset monitoring period, the diffusion rate of other edge points except the critical edge point is recorded. When the monitoring period expires, if the diffusion rate does not increase or the difference in diffusion rates is less than 0.5%, the system maintains the determination result that the lesion growth rate of the preset patient is slow; otherwise, it does not maintain.

[0021] As a preferred embodiment of the present invention, if the determination result is not maintained, the actual diffusion rate of each edge point except the critical edge point is calculated. Among the actual diffusion rates, 3 to 5 edge points with the highest diffusion rates are obtained, and these edge points are marked as risk edge points. The position distribution of the risk edge points is obtained. If the risk edge points are circularly distributed, the system determines that the area corresponding to the lesion location of the preset patient shows an overall diffusion trend; otherwise, it does not determine.

[0022] As a preferred embodiment of the present invention, if the system determines that the area corresponding to the lesion location of the preset patient shows an overall diffusion trend, the distance from the center to the edge of the risk edge point is calculated and marked as the reference distance. When the head CT image data of the preset patient is obtained in a future period, if the distance from the center to the edge of each edge point in the risk edge points is greater than the reference distance, and the distance difference is less than 0.5 cm, the system determines that the diffusion rate of the area corresponding to the lesion location of the preset patient is uniform and slow in multiple directions; otherwise, it does not determine.

[0023] As a preferred embodiment of the present invention, if it is not determined that the diffusion rate of the area corresponding to the lesion location of the preset patient is uniform and slow in multiple directions, the total distance differences obtained are statistically analyzed among the risk edge points. 10 of the largest distance differences are collected from all the distance differences, and the proportion of the 10 distance differences in all the distance differences is calculated. If the proportion in all the distance differences is less than one-half, the system determines that the diffusion trend of the area corresponding to the lesion location of the preset patient is local diffusion; otherwise, it does not determine.

[0024] Beneficial effects:

[0025] 1. By using methods such as wavelet transform, the present invention extracts the edge image features of the lesion, can more accurately identify the boundary of the lesion area, improves the diagnostic accuracy; and can quickly identify the lesion shape, location and density, and evaluate the lesion nature, reducing the time and workload of manual analysis.

[0026] 2. The present invention can analyze the change characteristics of the edge pixels in the lesion area based on CT image data at different times, and determine whether the lesion area shows a spreading trend; moreover, by calculating the time difference between the two obtained image data and the spreading rate of the lesion location, the progress speed of the lesion can be quantified, providing more detailed disease dynamics for doctors.

[0027] 3. The present invention identifies the edge point with the largest spreading area as the critical edge point, analyzes its spreading direction and characteristics, further evaluates the patient's symptom manifestations and disease progress, and by counting the proportion of the edge points with the slowest spreading rate, the system can determine whether the lesion speed slows down, providing a reference for clinical treatment.

[0028] 4. The present invention marks the edge point with the highest spreading rate as the risk edge point, analyzes its position distribution, and predicts the spreading trend of the lesion in advance. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following-described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings. Among them:

[0030] Figure 1 It is a modular structure schematic diagram of the intelligent analysis and diagnosis assistance system for medical image data according to the embodiment of the present invention;

[0031] Figure 2 It is a process structure schematic diagram according to the embodiment of the present invention;

[0032] Reference numerals in the figure: 110 - Image data acquisition module; 120 - Image data fusion processing unit; 1201 - Feature acquisition module; 1202 - Analysis module; 1203 - Comparison module; 1204 - Calculation module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions of the embodiments of the present invention in conjunction with the drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the described embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present invention.

[0034] In the head CT images of patients, the morphology, location, and density of brain lesions can vary greatly due to individual differences, making it difficult for AI models to accurately identify and classify them. As a result, the analysis accuracy of the AI model is reduced, and some early or subtle brain lesions (such as minor cerebral hemorrhage and early cerebral infarction) may be difficult to detect by existing AI models.

[0035] Based on this, the present invention proposes an intelligent analysis and diagnosis assistance system for medical image data. By means of automated analysis, dynamic monitoring, precise evaluation, prediction and early warning, and optimization of monitoring strategies, it improves the diagnostic efficiency and accuracy, provides comprehensive diagnostic support for doctors, helps improve the quality of medical services, and helps doctors formulate treatment plans in advance.

[0036] The following further specifically describes this solution through embodiments in conjunction with the accompanying drawings.

[0037] Refer to Figures 1 to 2 , which is an embodiment of the present invention. This embodiment provides an intelligent analysis and diagnosis assistance system for medical image data, including:

[0038] An image data acquisition module 110, which is used to acquire the head CT image data of a preset patient and mark the head CT image data as initial head CT image data; analyze the lesion morphology, lesion location, and lesion density of the patient's brain based on the head CT image data, and evaluate the nature of the lesion;

[0039] It should be noted in this embodiment that marking the initial CT image data as "initial head CT image data" facilitates subsequent comparative analysis with the image data at subsequent time points and provides a basis for dynamically monitoring the progression of lesions;

[0040] An image data fusion processing unit 120, which responds to the evaluation result of the nature of the lesion and is used to obtain the edge image features of the lesion morphology according to the image data; the image data fusion processing unit 120 includes a feature acquisition module 1201, an analysis module 1202, and a comparison module 1203;

[0041] The feature acquisition module 1201 is used to acquire the edge image features of the lesion morphology, and its acquisition method includes decomposing and reconstructing the edge image of the lesion morphology based on the wavelet transform method to extract the edge image features;

[0042] It should be noted in this embodiment that extracting the edge image features of the lesion morphology through methods such as wavelet transform can more accurately identify the boundary of the lesion area and improve the diagnostic accuracy;

[0043] At the same time, dynamically update the evaluation result of the nature of the lesion according to the change of the edge pixels to ensure the timeliness and accuracy of the diagnostic result;

[0044] In this embodiment, the wavelet transform is a mathematical tool used to decompose a signal or an image into basic functions (referred to as wavelet basis functions) at different scales and positions. It is a multi-resolution analysis method widely used in medical image processing, especially in edge extraction and feature extraction. Through multi-scale analysis, it can effectively capture the detailed information in the image and has good robustness to noise;

[0045] The analysis module 1202 responds to the extracted edge image features, sets a gray threshold according to the edge image features, and classifies the pixels in the edge image into edge pixels and non-edge pixels; marks the non-edge pixels as reference edge pixels; and updates the evaluation result of the lesion nature according to the change of the reference edge pixels;

[0046] The comparison module 1203 marks the head CT image data obtained for the preset patient in the future time period as the reference head CT image data; analyzes the change characteristics of the reference edge pixels based on the edge pixels of the reference head CT image data. If the change characteristics of the reference edge pixels are the same as those of the edge pixels, the system determines that the area corresponding to the lesion location of the preset patient shows a trend of diffusion change, and at the same time updates the evaluation result of the lesion nature to deterioration; otherwise, it does not determine;

[0047] In this embodiment, in the CT image, the edge pixels are the regions with significant gray changes in the image, and these regions usually correspond to the boundaries of tissues (such as the boundaries between the brain tissue and the ventricles, sulci, or the boundaries between the lesion area and the normal tissue); while the non-edge pixels refer to the regions with relatively small gray changes in the image, usually corresponding to the internal or background regions of the tissue;

[0048] The edge intensity / pixel (extracted by the wavelet transform method) and continuity of the lesion area can reflect the clarity of the lesion boundary. If the lesion edge becomes blurred or discontinuous, it may indicate that the lesion is spreading or deteriorating; for example, in the case of cerebral hemorrhage or cerebral edema, as the condition worsens, the boundary of the bleeding or edema area may gradually become blurred; by comparing CT images at different times, observing the changes in the lesion edge; if the area of the edge region increases significantly in a short period of time, it may indicate the deterioration of the condition; for example, the expansion rate of the perifocal edema (PHE) after cerebral hemorrhage is related to the deterioration of the condition;

[0049] Therefore, the determination method of this embodiment has practical significance;

[0050] It should be emphasized that the image data fusion processing unit 120 further includes a calculation module 1204. The calculation module 1204 is used to calculate the time difference between the acquisition of the reference head CT image data and the initial head CT image data based on the reference head CT image data and the initial head CT image data, and calculate the daily diffusion rate of the lesion location of the preset patient according to the time difference.

[0051] It should be noted in this embodiment that by calculating the time difference between the acquisition of the two image data and the diffusion rate of the lesion location, the progression speed of the lesion can be quantified, providing more detailed disease dynamics for doctors.

[0052] Moreover, it provides a quantitative index for subsequent dynamic monitoring, helping doctors better track the progression of the lesion.

[0053] Based on the above, in this embodiment, according to the calculated diffusion rate, the reference edge pixels are divided into edge points in the following way; where represents the nth edge point, and the diffusion characteristics of different edge points are analyzed within the time difference, and an edge point with the largest diffusion area is obtained, and the symptom manifestations and / or disease progression of the preset patient are analyzed according to the edge point.

[0054] In this embodiment, according to the research experience of the patient's head CT images, the continuity of the edge can reflect the integrity of the lesion area. If the edge appears broken or discontinuous, it may indicate the spread of the lesion area. Therefore, observing the change in the edge shape (such as whether it becomes irregular or serrated) can be associated with the patient's symptom manifestations and disease progression.

[0055] Furthermore, in this embodiment, an edge point with the largest diffusion area is marked as the critical edge point, the diffusion direction of the critical edge point is analyzed, and the diffusion rate of each other edge point is calculated according to the diffusion characteristics of the critical edge point within the time difference, and is calculated according to the following formula:

[0056] where p j represents the pth diffusion rate calculated for each edge point at the jth time point within the time difference;

[0057] In the formula, u represents the gray value of each edge point when the calculated pth diffusion rate is reached, t represents time, and the time is expressed as a percentage. This time percentage represents the percentage of the time point when the calculated pth diffusion rate is reached in the time difference; d is the diffusion coefficient, representing the diffusion rate; v 2 represents the second-order derivative in space;

[0058] It should be noted that in this embodiment, the edge point with the largest diffusion area is identified as the critical edge point, and its diffusion direction and characteristics are analyzed to further evaluate the patient's symptom manifestations and disease progression;

[0059] Moreover, through the analysis of the critical edge point, the key area of the lesion can be more accurately located, providing an important reference for subsequent treatment;

[0060] In this embodiment, according to the research experience of the patient's head CT images, by analyzing the direction of edge diffusion, it can be determined whether the lesion spreads along specific anatomical structures (such as cerebral sulci and gyri), which is beneficial to the study of the disease development;

[0061] At the same time, in the head CT image, if the diffusion rate of an edge point is known, the diffusion rates of other edge points can be inferred; for example, the diffusion rate of the hematoma edge of cerebral hemorrhage can be simulated and predicted;

[0062] On the above basis, according to the calculation results, the number of other edge points except the critical edge point is counted. Among the numbers, the 5-10 edge points with the slowest diffusion rate are counted according to the calculation results, and the proportion of the edge points in the number is calculated. If the proportion of the edge points exceeds half of the number, the system determines that the lesion speed of the preset patient is slow; otherwise, it is not determined.

[0063] It should be noted that in this embodiment, by counting the proportion of the edge points with the slowest diffusion rate, the system can determine whether the lesion speed slows down, providing a reference for clinical treatment, helping doctors judge the activity level of the lesion, and thus formulating a more appropriate treatment plan;

[0064] In this embodiment, in medical image analysis, the diffusion rate of edge points is related to the lesion speed, and it is necessary to conduct a comprehensive analysis in combination with specific lesion types, image characteristics, and clinical backgrounds. In some cases, the slow diffusion rate of edge points does reflect the slowdown of the lesion speed. For example, in cerebral edema, if the edge diffusion rate of the cerebral edema area slows down, it may indicate that the scope of the edema no longer expands and the condition tends to be stable; however, there are differences in the lesion characteristics and disease progression of different patients, and comprehensive judgment needs to be combined with clinical information.

[0065] Specifically in this embodiment, if the system determines that the lesion speed of the preset patient is slow, a monitoring period at least twice the time difference is preset; starting from the first day of the preset monitoring period, the diffusion rates of other edge points except the critical edge point are recorded. When the monitoring period expires, if the diffusion rate does not increase or the difference in the diffusion rate is less than 0.5%, the system maintains the determination result that the lesion speed of the preset patient is slow; otherwise, it does not maintain.

[0066] It should be noted that in this embodiment, if the system determines that the lesion progression slows down, it will automatically extend the monitoring period and continuously record the diffusion rate of the marginal points to ensure the continuity and accuracy of the disease condition monitoring;

[0067] And in a feasible solution, the monitoring strategy is dynamically adjusted according to the change of the lesion progression to avoid over-monitoring or under-monitoring;

[0068] On the above basis, if the determination result is not maintained, the actual diffusion rates of other marginal points except the critical marginal points are calculated, 3 to 5 marginal points with the highest diffusion rates are obtained from the actual diffusion rates, the marginal points are marked as risk marginal points, the position distribution of the risk marginal points is obtained. If the risk marginal points are distributed in a ring shape, the system determines that the area corresponding to the lesion position of the preset patient shows an overall diffusion trend; otherwise, no determination is made;

[0069] It should be noted that in this embodiment, by analyzing the distribution of the risk marginal points, the system can judge whether the lesion area shows an overall diffusion trend, which helps the doctor formulate a treatment plan in advance, and by analyzing the overall diffusion trend of the lesion area, it provides a more comprehensive basis for clinical decision-making;

[0070] In this embodiment, in medical image analysis, when the diffusion of the marginal points shows a ring-shaped distribution, it usually means that the boundary of the lesion area expands uniformly in multiple directions; for example, the marginal diffusion of the brain edema area may show a ring-shaped distribution, especially around the ventricles or near the cerebral sulci. This diffusion pattern may indicate that the edema area expands uniformly in multiple directions, resulting in a gradual increase in the area of the lesion area;

[0071] Among them, if the system determines that the area corresponding to the lesion position of the preset patient shows an overall diffusion trend, the distance from the center to the edge of the risk marginal points is calculated and marked as the reference distance; when the head CT image data of the preset patient is obtained in the future time period, if the distances from the center to the edge of each marginal point in the risk marginal points are all greater than the reference distance, but the distance differences are all less than 0.5 cm, the system determines that the diffusion rate of the area corresponding to the lesion position of the preset patient is uniform and slow in multiple directions; otherwise, no determination is made;

[0072] In this embodiment, if the diffusion rate of the marginal points is uniform and slow in multiple directions, it indicates that the overall diffusion trend of the lesion area is relatively stable; otherwise, it indicates that the overall diffusion trend of the lesion area is relatively rapid;

[0073] Furthermore, if it is not determined that the diffusion rate of the area corresponding to the lesion location of the preset patient is uniform and slow in multiple directions, then all the obtained distance differences are statistically analyzed among the risk marginal points, 10 largest distance differences are collected from all the distance differences, and the proportion of the 10 distance differences in all the distance differences is calculated. If the proportion in all the distance differences is less than one half, the system determines that the diffusion trend of the area corresponding to the lesion location of the preset patient is local diffusion; otherwise, it is not determined.

[0074] It should be noted in this embodiment that by statistically analyzing the distance differences, the system can more precisely evaluate the diffusion trend of the lesion and help doctors formulate more accurate treatment plans.

[0075] In summary, through automated analysis, dynamic monitoring, precise evaluation, prediction and early warning, and optimized monitoring strategies, the present invention improves the diagnostic efficiency and accuracy, provides comprehensive diagnostic support for doctors, helps improve the quality of medical services, and helps doctors formulate treatment plans in advance.

[0076] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not 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. An intelligent analysis and diagnosis assistance system for medical imaging data, characterized in that, Including: An image data acquisition module, which is used to acquire the head CT image data of a preset patient, mark the head CT image data as initial head CT image data; analyze the lesion shape, lesion location and lesion density of the patient's brain based on the head CT image data, and evaluate the lesion nature; An image data fusion processing unit, which responds to the evaluation result of the lesion nature and is used to obtain the edge image features of the lesion shape according to the image data; the image data fusion processing unit includes a feature acquisition module, an analysis module and a comparison module; The feature acquisition module is used to acquire the edge image features of the lesion shape, and its acquisition method includes decomposing and reconstructing the edge image of the lesion shape based on the wavelet transform method, so as to extract the edge image features; The analysis module responds to the extracted edge image features, and is used to set a gray threshold according to the edge image features, and distinguish the pixel areas in the edge image into edge pixels and non-edge pixels; mark the non-edge pixels as reference edge pixels; and update the evaluation result of the lesion nature according to the change of the reference edge pixels; The comparison module marks the head CT image data obtained for the preset patient in the future time period as reference head CT image data; it is used to analyze the change characteristics of the reference edge pixels based on the edge pixels of the reference head CT image data. If the change characteristics of the reference edge pixels are the same as those of the edge pixels, the system determines that the area corresponding to the lesion location of the preset patient shows a diffusion trend change, and at the same time updates the evaluation result of the lesion nature to deterioration; Otherwise, no determination is made.

2. The intelligent analysis and diagnosis assistance system for medical imaging data according to claim 1, wherein The image data fusion processing unit further includes a calculation module, which is used to calculate the time difference between the acquisitions of the head CT image data twice according to the reference head CT image data and the initial head CT image data, and calculate the daily diffusion rate of the lesion location of the preset patient according to the time difference.

3. The intelligent analysis and diagnosis assistance system for medical image data according to claim 2, characterized in that, According to the calculated diffusion rate, the reference edge pixels are divided into edge points in the manner of ; where represents the nth edge point, and the diffusion characteristics of different edge points are analyzed within the time difference, an edge point with the largest diffusion area is obtained, and the symptom manifestations and / or disease progression of a preset patient are analyzed according to the edge point.

4. An intelligent analysis and diagnosis assistance system for medical imaging data according to claim 3, characterized in that, Mark the edge point with the largest diffusion area as the critical edge point, analyze the diffusion direction of the critical edge point, and calculate the diffusion rate of other edge points according to the diffusion characteristics of the critical edge point within the time difference, which is calculated according to the following formula: where p j represents the p-th diffusion rate calculated for each edge point at the j-th time point within the time difference; Wherein, u represents the gray value of each edge point up to the calculated p-th diffusion rate, t represents time, and the time is expressed as a percentage, and this time percentage represents the percentage of the time point up to the calculated p-th diffusion rate in the time difference; d is the diffusion coefficient, representing the diffusion rate; v 2 represents the second-order derivative in space.

5. An intelligent analysis and diagnosis assistance system for medical image data according to claim 4, characterized in that, According to the calculation results, count the number of other edge points except the critical edge point. Among the number, count 5-10 edge points with the slowest diffusion rate according to the calculation results, and calculate the proportion of the edge points in the number. If the proportion of the edge points exceeds half of the number, the system determines that the lesion speed of the preset patient is slow; Otherwise, no determination is made.

6. The intelligent analysis and diagnosis assistance system for medical image data according to claim 5, characterized in that, If the system determines that the lesion speed of the preset patient is slow, a monitoring period at least twice the time difference is preset; starting from the first day of the preset monitoring period, record the diffusion rate of other edge points except the critical edge point. When the monitoring period expires, if the diffusion rate does not increase or the difference in the diffusion rate is less than 0.5%, the system maintains the determination result that the lesion speed of the preset patient is slow; otherwise, it does not maintain.

7. An intelligent analysis and diagnosis assistance system for medical imaging data according to claim 6, characterized in that, If the determination result is not maintained, calculate the actual diffusion rate of each edge point other than the critical edge point, obtain 3 to 5 edge points with the highest diffusion rate from the actual diffusion rates, mark the edge points as risk edge points, obtain the position distribution of the risk edge points. If the risk edge points are distributed in a ring shape, the system determines that the area corresponding to the lesion position of the preset patient shows an overall diffusion trend; Otherwise, no determination is made.

8. An intelligent analysis and diagnosis assistance system for medical imaging data according to claim 7, characterized in that, If the system determines that the area corresponding to the lesion position of the preset patient shows an overall diffusion trend, calculate the distance from the center to the edge of the risk edge point and mark the distance as the reference distance. When obtaining the head CT image data of the preset patient in the future time period, if the distance from the center to the edge of each edge point in the risk edge points is greater than the reference distance, but the distance difference is less than 0.5 cm, the system determines that the diffusion rate of the area corresponding to the lesion position of the preset patient is uniform and slow in multiple directions; otherwise, no determination is made.

9. An intelligent analysis and diagnosis assistance system for medical imaging data as described in claim 8, characterized in that, If it is not determined that the diffusion rate of the area corresponding to the lesion position of the preset patient is uniform and slow in multiple directions, count all the obtained distance differences among the risk edge points, collect 10 largest distance differences from all the distance differences, calculate the proportion of the 10 distance differences in all the distance differences. If the proportion in all the distance differences is less than one-half, the system determines that the diffusion trend of the area corresponding to the lesion position of the preset patient is local diffusion; Otherwise, no determination is made.

Images