Medical image diagnosis auxiliary system based on animation design

Abstract

The invention relates to the technical field of medical images, and discloses a medical image diagnosis auxiliary system based on animation design, which comprises a data acquisition module, an image segmentation analysis module, an artifact analysis module, a three-dimensional model analysis module and a visual operation module. And full-link optimization of medical image processing is realized. By introducing an artifact identification coefficient AIC and a three-dimensional modeling analysis coefficient MRC, three-dimensional reconstruction is more accurate, and interference of image artifacts on diagnosis is reduced. Besides, an interactive visual interface improves the ability of doctors to understand and analyze lesions, more reliable data support is provided for surgical planning, and the image segmentation qualification identification coefficient TFX is calculated through multi-feature fusion, so that the method can more comprehensively measure the segmentation quality and reduce misjudgment and missed judgment, thereby ensuring the stability and reliability of the segmentation result, and improving the accuracy of surgical operation. And the practical application value of the system is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical imaging, and particularly to a medical imaging diagnosis assistance system based on animation design. Background Art

[0002] The medical imaging diagnosis assistance system belongs to the cross - technology field of artificial intelligence and healthcare. With the development of medical imaging equipment, especially the wide application of high - end imaging technologies such as CT and MRI, medical imaging has become an indispensable tool for clinical diagnosis.

[0003] Although medical imaging technology is becoming increasingly mature, it still faces many challenges in practical applications. Specifically, existing medical imaging processing systems often focus on solving single tasks, such as image segmentation or artifact removal, resulting in insufficient overall performance. First, there are still great challenges in the accuracy of image segmentation. Especially for complex lesion areas, existing algorithms often cannot accurately locate the boundaries of each lesion, and segmentation errors are likely to occur. Second, the presence of artifacts interferes with the true reflection of the image. In the case of complex artifact types, the existing technology still has limited ability in intelligent recognition and removal, resulting in ineffective removal of artifacts, which further affects the doctor's diagnostic judgment. Although three - dimensional modeling technology has made some progress, due to limitations in processing speed and accuracy, the generated three - dimensional models are still difficult to meet the requirements of clinical precision treatment and surgical planning.

[0004] Therefore, we propose a medical imaging diagnosis assistance system based on animation design to solve the above - mentioned problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a medical imaging diagnosis assistance system based on animation design to solve the problems in the above - mentioned background art that there are still great challenges in the accuracy of image segmentation. Especially for complex lesion areas, existing algorithms often cannot accurately locate the boundaries of each lesion, and segmentation errors are likely to occur. Second, the presence of artifacts interferes with the true reflection of the image. In the case of complex artifact types, the existing technology still has limited ability in intelligent recognition and removal, resulting in ineffective removal of artifacts, which further affects the doctor's diagnostic judgment. Although three - dimensional modeling technology has made some progress, due to limitations in processing speed and accuracy, the generated three - dimensional models are still difficult to meet the requirements of clinical precision treatment and surgical planning.

[0006] To achieve the above purpose, the present invention provides the following technical solution: A medical imaging diagnosis assistance system based on animation design, including a data acquisition module, an image segmentation and analysis module, an artifact analysis module, a three - dimensional model analysis module, and a visualization operation module;

[0007] The data acquisition module is used to extract data from medical images and preprocess the extracted data, so as to organize them into a first data set, a second data set, and a third data set;

[0008] The image segmentation analysis module is used to analyze the first data set, so as to generate an image segmentation qualified identification coefficient TFX, and analyze the image segmentation qualified identification coefficient TFX, so as to generate a first analysis result;

[0009] The artifact analysis module is used to analyze the second data set, so as to generate an artifact identification coefficient AIC, and analyze the artifact identification coefficient AIC, so as to generate a second analysis result;

[0010] The three-dimensional model analysis module is used to integrally analyze the third data set and the artifact identification coefficient AIC, so as to generate a three-dimensional modeling analysis coefficient MRC, and analyze the three-dimensional modeling analysis coefficient MRC, so as to generate a third analysis result;

[0011] The visualization operation module is used to project the three-dimensional model onto the visualization operation interface.

[0012] Preferably, the data acquisition module includes a data extraction unit and a data preprocessing unit;

[0013] Among them, the data extraction unit is used to perform multi-data acquisition on the image through acquisition software, including pixel resolution, slice thickness, contrast, gradient change rate, signal-to-noise ratio, gray-level co-occurrence matrix features, edge sharpness, tissue gray-level uniformity, scan time, scan energy, signal drift, stripe artifact intensity, metal artifact ratio, motion artifact amplitude, signal non-uniformity, scan angle error, voxel size, point cloud density, surface smoothness, topological connectivity, volume deviation, surface texture consistency, structure noise ratio, and model simplification error;

[0014] The data preprocessing unit preprocesses and dimensionlessizes the collected multiple data, and reorganizes them into a first data set, a second data set, and a third data set;

[0015] The first data set includes pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix features HG, edge sharpness RD, and tissue gray-level uniformity ZH;

[0016] The second data set includes scan time SS, scan energy SN, signal drift XP, stripe artifact intensity TW, metal artifact ratio JS, motion artifact amplitude YD, signal non-uniformity XB, and scan angle error WC;

[0017] The third data set includes voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structure noise ratio ZB, and model simplification error MX.

[0018] Preferably, the image segmentation analysis module includes an image segmentation identification unit and an image segmentation analysis unit;

[0019] The image segmentation identification unit is used to perform integrated calculations on the first data set. By integrating pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix feature HG, edge sharpness RD, and tissue gray-level uniformity ZH, the qualified image segmentation identification coefficient TFX is obtained.

[0020] The image segmentation analysis unit is used to compare the qualified image segmentation identification coefficient TFX with a preset image segmentation identification threshold Y1 to generate a first analysis result.

[0021] Preferably, the image segmentation identification unit calculates and obtains the qualified image segmentation identification coefficient TFX through the following formula:

[0022]

[0023] In the formula: FB is the pixel resolution, HD is the slice thickness, DB is the contrast, TD is the gradient change rate, XZ is the signal-to-noise ratio, HG is the gray-level co-occurrence matrix feature, RD is the edge sharpness, ZH is the tissue gray-level uniformity, and e is the base of the natural logarithm.

[0024] Preferably, the first analysis result generated by the image segmentation analysis unit is as follows:

[0025] When AIC > Y1, it means that the current image segmentation area is abnormal and needs to be recalibrated;

[0026] When AIC ≤ Y1, it means that the current image segmentation area is normal and proceed to the next step.

[0027] Preferably, the artifact analysis module includes an artifact identification unit and an artifact analysis unit;

[0028] The artifact identification unit is used to calculate the second data set. By integrating scan time SS, scan energy SN, signal drift XP, streak artifact intensity TW, metal artifact ratio JS, motion artifact amplitude YD, signal inhomogeneity XB, and scan angle error WC, the artifact identification coefficient AIC is obtained.

[0029] The artifact analysis unit is used to compare the calculated artifact identification coefficient AIC with a preset artifact identification threshold Y2 to generate a second analysis result.

[0030] Preferably, the artifact identification unit calculates and obtains an artifact identification coefficient AIC through the following formula:

[0031]

[0032] In the formula: SS is the scanning time, SN is the scanning energy, XP is the signal drift, TW is the stripe artifact intensity, JS is the metal artifact ratio, YD is the motion artifact amplitude, XB is the signal non-uniformity, WC is the scanning angle error, a is an adjustment coefficient, and the value range of a is 0.52 to 2, log is the logarithmic function, and tan -1 is the arctangent function.

[0033] Preferably, the artifact analysis unit generates a second analysis result, specifically as follows:

[0034] When AIC > Y2, it means that there are artifacts in the current image and secondary noise reduction is required;

[0035] When MRC ≤ Y2, it means that there are no artifacts in the current image and 3D modeling is performed.

[0036] Preferably, the 3D model analysis module includes a model identification calculation unit and a model analysis unit;

[0037] The model identification calculation unit is used to perform integrated calculation on the third data set and the artifact identification coefficient AIC, and obtains a 3D modeling analysis coefficient MRC by integrating the voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structural noise ratio ZB, model simplification error MX, and the artifact identification coefficient AIC;

[0038] The model analysis unit is used to compare the calculated 3D modeling analysis coefficient MRC with a preset 3D modeling qualification threshold Y3, so as to generate a third analysis result, specifically as follows;

[0039] When MRC > Y3, it means that the current 3D modeling meets the usage requirements and can be sent to the visualization terminal;

[0040] When MRC ≤ Y3, it means that the current 3D modeling does not meet the usage requirements and can be sent to the visualization terminal.

[0041] Preferably, the model identification calculation unit calculates and obtains a 3D modeling analysis coefficient MRC through the following formula:

[0042]

[0043] Where: TS is the voxel size, DY is the point cloud density, QM is the surface smoothness, TP is the topological connectivity, TC is the volume deviation, BM is the surface texture consistency, ZB is the structural noise ratio, MX is the model simplification error, ln is the natural logarithm, and tan -1 is the arctangent function, and e is the base of the natural logarithm.

[0044] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0045] 1. Through the mutual cooperation of multiple modules, the present system realizes the full-link optimization of medical image processing. The introduction of the artifact identification coefficient AIC and the three-dimensional modeling analysis coefficient MRC makes the three-dimensional reconstruction more accurate and reduces the interference of image artifacts on diagnosis. In addition, the interactive visualization interface improves the doctor's understanding and analysis ability of the lesion, providing more reliable data support for surgical planning.

[0046] 2. By calculating the image segmentation qualified identification coefficient TFX through multi-feature fusion, this method can measure the segmentation quality more comprehensively, reduce misjudgment and missed judgment, thereby ensuring the stability and reliability of the segmentation result, improving the practical application value of the system. Combining parameters such as the gradient change rate TD, the edge sharpness RD, and the tissue gray level uniformity ZH, this method can effectively identify complex tissue structures in regions with blurred boundaries and low contrast, thereby improving the ability of the segmentation algorithm to depict details and meeting the requirements of high-precision image analysis. Using mathematical formulas to quantitatively evaluate the quality of image segmentation, replacing the traditional quality judgment method relying on manual experience, improving the automation degree of the evaluation process, reducing human intervention, and thus improving the efficiency of large-scale image processing, which is applicable to real-time monitoring and intelligent image analysis scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is the system step diagram of the present invention.

[0048] In the figure: 1. Data acquisition module; 11. Data extraction unit; 12. Data preprocessing unit; 2. Image segmentation and analysis module; 21. Image segmentation identification unit; 22. Image segmentation analysis unit; 3. Artifact analysis module; 31. Artifact identification unit; 32. Artifact analysis unit; 4. Three-dimensional model analysis module; 41. Model identification and calculation unit; 42. Model analysis unit; 5. Visualization operation module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0050] Embodiment 1: Please refer to Figure 1 , a medical image diagnosis assistance system based on animation design, including a data acquisition module 1, an image segmentation and analysis module 2, an artifact analysis module 3, a three-dimensional model analysis module 4, and a visualization operation module 5;

[0051] The data acquisition module 1 is used to extract data from medical images and preprocess the extracted data, so as to organize it into a first data set, a second data set, and a third data set;

[0052] The image segmentation and analysis module 2 is used to analyze the first data set, so as to generate an image segmentation qualified identification coefficient TFX, and analyze the image segmentation qualified identification coefficient TFX, so as to generate a first analysis result;

[0053] The artifact analysis module 3 is used to analyze the second data set, so as to generate an artifact identification coefficient AIC, and analyze the artifact identification coefficient AIC, so as to generate a second analysis result;

[0054] The three-dimensional model analysis module 4 is used to integrally analyze the third data set and the artifact identification coefficient AIC, so as to generate a three-dimensional modeling analysis coefficient MRC, and analyze the three-dimensional modeling analysis coefficient MRC, so as to generate a third analysis result;

[0055] The visualization operation module 5 is used to project the three-dimensional model onto the visualization operation interface.

[0056] In this embodiment: The data acquisition module 1 is responsible for the acquisition, preprocessing, and classification and sorting of medical images to ensure the integrity and accuracy of the data. Through advanced data extraction technology, this module can effectively filter noise, enhance the image quality, and classify the information into the first data set image feature data, the second data set artifact feature data, and the third data set three-dimensional modeling feature data. This process improves the reliability of subsequent analysis and provides a solid data foundation for accurate diagnosis.

[0057] The image segmentation analysis module 2 uses the image segmentation qualified identification coefficient TFX to evaluate the segmentation effect of medical images, ensuring that the anatomical structures are clearly distinguishable. Through deep learning algorithms and edge detection techniques, this module can effectively distinguish the lesion area from normal tissues, improving the segmentation accuracy. Finally, this module analyzes TFX, outputs the first analysis result, optimizes the subsequent processing of medical images, improves the doctor's ability to identify lesions, and enhances the accuracy and stability of auxiliary diagnosis.

[0058] The artifact analysis module 3 analyzes the second data set, calculates the artifact identification coefficient AIC, and identifies and quantifies the impact of artifacts on medical images. Since artifacts may lead to misdiagnosis or reduce image clarity, the role of this module is to detect artifact types such as metal artifacts, motion artifacts, signal drift, etc. and analyze their impact degree, and finally generate the second analysis result. This process helps to optimize the quality of medical images and improve the doctor's diagnostic accuracy under low artifact interference.

[0059] The 3D model analysis module 4 fuses the third data set with the artifact identification coefficient AIC, calculates the 3D modeling analysis coefficient MRC, and evaluates and optimizes the 3D reconstruction quality of medical images. This module not only considers the core 3D modeling parameters such as voxel size, point cloud density, and topological connectivity, but also combines the impact of artifacts on modeling to ensure that the finally generated 3D model has both structural integrity and can minimize artifact interference to the greatest extent. Finally, this module analyzes MRC, generates the third analysis result, improves the usability of 3D medical modeling in clinical applications, enables doctors to more intuitively observe the lesion morphology, and improves the accuracy of surgical planning and preoperative evaluation.

[0060] The visualization operation module 5 projects the optimized 3D model onto the visualization interface, providing an interactive display of medical images. This module supports operations such as rotation, scaling, and layer display, enabling doctors to analyze lesions from multiple angles and improving the intuitiveness of image diagnosis. At the same time, this module can also be used in application scenarios such as surgical navigation, preoperative planning, and medical teaching, improving the application value of medical images and promoting the development of precision medicine.

[0061] Through the mutual cooperation of multiple modules, this system realizes the full-link optimization of medical image processing. The introduction of the artifact identification coefficient AIC and the 3D modeling analysis coefficient MRC makes the 3D reconstruction more accurate and reduces the interference of image artifacts on diagnosis. In addition, the interactive visualization interface improves the doctor's understanding and analysis ability of lesions, providing more reliable data support for surgical planning.

[0062] Example 2: Please refer to Figure 1 , the data acquisition module 1 includes a data extraction unit 11 and a data preprocessing unit 12;

[0063] Among them, the data extraction unit 11 is used to perform multi-data acquisition on the image through acquisition software, including pixel resolution, slice thickness, contrast, gradient change rate, signal-to-noise ratio, gray-level co-occurrence matrix features, edge sharpness, tissue gray-level uniformity, scan time, scan energy, signal drift, stripe artifact intensity, metal artifact ratio, motion artifact amplitude, signal non-uniformity, scan angle error, voxel size, point cloud density, surface smoothness, topological connectivity, volume deviation, surface texture consistency, structure noise ratio, and model simplification error;

[0064] The data preprocessing unit 12 preprocesses and dimensionlessizes the collected multiple data, and reorganizes them into a first data set, a second data set, and a third data set;

[0065] The first data set includes pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix features HG, edge sharpness RD, and tissue gray-level uniformity ZH;

[0066] The second data set includes scan time SS, scan energy SN, signal drift XP, stripe artifact intensity TW, metal artifact ratio JS, motion artifact amplitude YD, signal non-uniformity XB, and scan angle error WC;

[0067] The third data set includes voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structure noise ratio ZB, and model simplification error MX.

[0068] In this embodiment: The data extraction unit 11 of this system not only acquires traditional medical image parameters such as pixel resolution, slice thickness, and contrast, but also additionally introduces advanced features such as gradient change rate, gray-level co-occurrence matrix features, edge sharpness, and tissue gray-level uniformity. These parameters can comprehensively characterize the image quality and structural details, helping the system accurately evaluate the clarity, contrast, and edge information of the image, thereby optimizing image segmentation and enhancing the diagnostic value of medical images.

[0069] The data preprocessing unit 12 performs standardization processing and dimensionless conversion on the collected data to ensure the numerical comparability between different physical quantities, and improves the consistency and calculation efficiency of the data. Through reasonable normalization processing, this module can eliminate the influence of different units and scales, making the analysis of the first data set, the second data set, and the third data set more accurate, and providing a more stable data input for subsequent algorithm processing and analysis.

[0070] Traditional imaging systems often only focus on image clarity while ignoring the impact of artifacts on image quality. The second dataset of this system is specifically for artifact analysis, covering various artifact features such as scanning time, scanning energy, signal drift, stripe artifact intensity, metal artifact ratio, motion artifact amplitude, signal non-uniformity, and scanning angle error. In-depth analysis of these data can accurately identify the types and severity of artifacts, providing an accurate basis for the calculation of the subsequent artifact identification coefficient AIC, thereby reducing the interference of artifacts on diagnosis and enhancing the reliability of medical images.

[0071] The third dataset integrates key parameters for 3D modeling, such as voxel size, point cloud density, surface smoothness, topological connectivity, volume deviation, surface texture consistency, structure noise ratio, and model simplification error. These metrics can comprehensively evaluate the fineness and topological structure of 3D models. Compared with traditional imaging modeling methods, this system can optimize the surface quality of the model, reduce noise interference, and improve the accuracy of geometric structures, providing accurate data support for subsequent 3D modeling analysis, thereby generating high-quality models that better meet clinical needs.

[0072] Embodiment 3: Please refer to Figure 1 , the image segmentation analysis module 2 includes an image segmentation identification unit 21 and an image segmentation analysis unit 22;

[0073] The image segmentation identification unit 21 is used to perform integrated calculations on the first dataset. By integrating and calculating the pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix feature HG, edge sharpness RD, and tissue gray-level uniformity ZH, the image segmentation qualified identification coefficient TFX is obtained.

[0074] The image segmentation analysis unit 22 is used to compare the image segmentation qualified identification coefficient TFX with a preset image segmentation identification threshold Y1 to generate a first analysis result.

[0075] The image segmentation identification unit 21 calculates and obtains the image segmentation qualified identification coefficient TFX through the following formula:

[0076]

[0077] In the formula: FB is the pixel resolution, HD is the slice thickness, DB is the contrast, TD is the gradient change rate, XZ is the signal-to-noise ratio, HG is the gray-level co-occurrence matrix feature, RD is the edge sharpness, ZH is the tissue gray-level uniformity, and e is the base of the natural logarithm.

[0078] The first analysis result generated by the image segmentation analysis unit 22 is as follows:

[0079] When AIC > Y1, it means that the current image segmentation area is abnormal and needs to be recalibrated;

[0080] When AIC ≤ Y1, it indicates that the current image segmentation region is normal, and the next step of processing is carried out.

[0081] In this embodiment: The image segmentation recognition unit 21 calculates the image segmentation qualified recognition coefficient TFX by integrating the pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix feature HG, edge sharpness RD, and tissue gray-level uniformity ZH. These parameters not only cover the basic quality indicators of the image but also combine advanced imaging features such as gradient change rate, edge sharpness, and signal-to-noise ratio, so that the calculated TFX can accurately reflect the quality of image segmentation, thus ensuring more accurate segmentation of the lesion area and avoiding the situation of mis-segmentation or over-segmentation.

[0082] The formula comprehensively considers multiple key features such as pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix feature HG, edge sharpness RD, and tissue gray-level uniformity ZH to more comprehensively evaluate the quality of image segmentation. Compared with the traditional evaluation method that relies on a single feature, this method can effectively reduce one-sidedness, improve the accuracy and stability of the segmentation quality evaluation, uses non-linear transformation to weight each feature, enabling the system to adaptively adjust the calculation weights according to the characteristics of different images, so as to accurately evaluate the quality of image segmentation under different imaging conditions and tissue structures, enhance the generalization ability and robustness of the system. By introducing the gray-level co-occurrence matrix feature HG, this method can more deeply analyze the spatial relationship between pixels, thereby enhancing the ability to extract tissue detail features. Especially in the regions with blurred boundaries and low contrast, it can effectively improve the refinement degree of the segmentation quality evaluation, and is applicable to high-precision application scenarios such as medical image processing. Since the slice thickness HD has an important impact on the segmentation accuracy of three-dimensional imaging such as CT and MRI, this method fully considers the slice thickness factor when calculating the image segmentation quality to ensure that the evaluation standard is more in line with the actual application requirements, thereby improving the evaluation accuracy of three-dimensional image segmentation and enhancing the applicability of the system in three-dimensional medical image analysis.

[0083] By calculating the qualified identification coefficient TFX of image segmentation through multi-feature fusion, this method can measure the segmentation quality more comprehensively, reduce misjudgment and missed judgment, thereby ensuring the stability and reliability of the segmentation result, improving the practical application value of the system. Combining parameters such as the gradient change rate TD, edge sharpness RD, and tissue gray uniformity ZH, this method can effectively identify complex tissue structures in regions with blurred boundaries and low contrast, thereby improving the ability of the segmentation algorithm to depict details and meeting the requirements of high-precision image analysis. Using mathematical formulas to quantitatively evaluate the quality of image segmentation, replacing the traditional quality judgment method relying on manual experience, improving the automation degree of the evaluation process, reducing human intervention, and thus enhancing the efficiency of large-scale image processing, which is applicable to real-time monitoring and intelligent image analysis scenarios.

[0084] The image segmentation analysis unit 22 introduces a standardized automatic evaluation mechanism, compares the calculated TFX with the preset image segmentation identification threshold Y1, and automatically generates the first analysis result. This method can avoid human errors, improve the objectivity of the segmentation result, and ensure that all images meet the set quality standards, thereby enhancing the stability and consistency of image segmentation.

[0085] By setting the key parameter AIC segmentation region anomaly index to compare with the threshold Y1, this method can achieve intelligent judgment of the quality of image segmentation, ensuring the accuracy and stability of the evaluation process. Compared with the traditional quality determination method relying on manual experience, this method reduces subjective errors, improves the automation degree and consistency of the system. By setting that when AIC > Y1, the recalibration mechanism is triggered, this method can automatically detect abnormal situations in the segmentation region and make timely adjustments to reduce mis-segmentation, over-segmentation or under-segmentation, ensuring that the final segmentation result meets the expected standard and improving the segmentation accuracy. By setting that when AIC ≤ Y1, it directly enters the next processing step, this method can quickly identify the segmentation regions that meet the quality standards, avoid unnecessary repeated calculations, thereby improving the processing efficiency and reducing the consumption of system resources, making it more suitable for large-scale image processing tasks. Using the dynamic threshold Y1 to evaluate the quality of the segmentation region enables this method to adapt to different imaging conditions, tissue types and application scenarios, ensuring reliable quality evaluation on different image data sets and improving the adaptability and generalization ability of the algorithm.

[0086] Example 4: Please refer to Figure 1 , the artifact analysis module 3 includes an artifact identification unit 31 and an artifact analysis unit 32;

[0087] The artifact identification unit 31 is used to calculate the second data set, and obtains the artifact identification coefficient AIC by integrating the scanning time SS, the scanning energy SN, the signal drift XP, the streak artifact intensity TW, the metal artifact ratio JS, the motion artifact amplitude YD, the signal non-uniformity XB, and the scanning angle error WC;

[0088] The artifact analysis unit 32 is used to compare the calculated artifact identification coefficient AIC with a preset artifact identification threshold Y2, so as to generate a second analysis result.

[0089] The artifact identification unit 31 calculates and obtains the artifact identification coefficient AIC through the following formula:

[0090]

[0091] In the formula: SS is the scanning time, SN is the scanning energy, XP is the signal drift, TW is the streak artifact intensity, JS is the metal artifact ratio, YD is the motion artifact amplitude, XB is the signal non-uniformity, WC is the scanning angle error, a is an adjustment coefficient, and the value range of a is 0.52 to 2, log is the logarithmic function, and tan -1 is the arctangent function.

[0092] The artifact analysis unit 32 generates a second analysis result, specifically as follows:

[0093] When AIC > Y2, it means that there are artifacts in the current image and secondary noise reduction is required;

[0094] When MRC ≤ Y2, it means that there are no artifacts in the current image and 3D modeling is performed.

[0095] In this embodiment: The artifact identification unit 31 calculates the artifact identification coefficient AIC by integrating multiple key parameters such as the scanning time SS, the scanning energy SN, the signal drift XP, the streak artifact intensity TW, the metal artifact ratio JS, the motion artifact amplitude YD, the signal non-uniformity XB, and the scanning angle error WC. Compared with the traditional artifact detection method that only relies on individual factors, this method can analyze the factors affecting artifact generation more comprehensively, improve the accuracy of artifact recognition, reduce misjudgment and missed judgment. Through the combined operation of exponential functions, logarithmic functions, and arctangent functions, this method can dynamically adjust the calculation weight according to the influence degree of different artifact characteristics, ensuring that AIC can effectively measure the artifact interference level in different environments, thereby enhancing the generalization ability of the system under different scanning conditions.

[0096] The artifact analysis unit 32 automatically determines whether the current image requires secondary noise reduction or directly enters the 3D modeling stage by comparing the AIC with the preset artifact identification threshold Y2. This method avoids the subjective errors of traditional manual interpretation, improves the intelligence level of the system, and makes the artifact processing process more efficient. When AIC > Y2, the system automatically triggers secondary noise reduction, effectively reducing image artifact problems caused by motion artifacts, metal artifacts, signal drift, etc., and improving the readability and diagnostic accuracy of medical images. Compared with the fixed noise reduction strategy, this method can perform targeted processing according to the degree of artifacts, avoid unnecessary noise reduction processes, improve the imaging quality while reducing information loss.

[0097] Example 5: Please refer to Figure 1 , the 3D model analysis module 4 includes a model identification calculation unit 41 and a model analysis unit 42;

[0098] The model identification calculation unit 41 is used to perform integrated calculations on the third data set and the artifact identification coefficient AIC, and obtain the 3D modeling analysis coefficient MRC by integrating the voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structural noise ratio ZB, model simplification error MX, and the artifact identification coefficient AIC;

[0099] The model analysis unit 42 is used to compare the calculated 3D modeling analysis coefficient MRC with the preset 3D modeling qualification threshold Y3 to generate a third analysis result, specifically as follows;

[0100] When MRC > Y3, it means that the current 3D modeling meets the usage requirements and can be sent to the visualization terminal;

[0101] When MRC ≤ Y3, it means that the current 3D modeling does not meet the usage requirements and can be sent to the visualization terminal.

[0102] The model identification calculation unit 41 calculates and obtains the 3D modeling analysis coefficient MRC through the following formula:

[0103]

[0104] In the formula: TS is the voxel size, DY is the point cloud density, QM is the surface smoothness, TP is the topological connectivity, TC is the volume deviation, BM is the surface texture consistency, ZB is the structural noise ratio, MX is the model simplification error, ln is the natural logarithm, tan -1 is the arctangent function, and e is the base of the natural logarithm.

[0105] In this embodiment: The model identification calculation unit 41 integrates and calculates multiple key features such as voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structural noise ratio ZB, model simplification error MX, and artifact identification coefficient AIC, so as to generate a three-dimensional modeling analysis coefficient MRC. The fusion of such multi-dimensional features can more comprehensively evaluate the quality of the three-dimensional model, avoid the deficiencies of single-feature evaluation, and enhance the accuracy and reliability of model evaluation.

[0106] By introducing various mathematical operations such as the natural logarithm function ln, exponential function e, and inverse tangent function tan-1 into the MRC calculation formula, this method can dynamically adjust weights and calculation methods according to different modeling characteristics, enabling the system to adapt to different types of three-dimensional modeling requirements, and improving the flexibility and robustness of the algorithm.

[0107] During the three-dimensional modeling analysis process, factors affecting the artifact identification coefficient AIC, such as artifact intensity, metal artifacts, and motion artifacts, are considered, which can accurately evaluate the interference of artifacts on the modeling effect, ensuring that the three-dimensional modeling result can still reflect the real physical structure in the presence of artifacts. This effectively reduces the impact of artifacts on the model quality, making the final three-dimensional model more accurate. By comparing with the preset three-dimensional modeling qualification threshold Y3, the model analysis unit 42 can automatically determine whether the three-dimensional model meets the usage requirements and make corresponding processing. This automated evaluation mechanism reduces manual intervention, improves work efficiency, and ensures the consistency of the discrimination criteria, making the image analysis process more efficient and accurate.

[0108] The content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0109] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A medical image diagnosis assistance system based on animation design, characterized in that: It includes a data acquisition module (1), an image segmentation and analysis module (2), an artifact analysis module (3), a 3D model analysis module (4), and a visualization operation module (5); The data acquisition module (1) is used to extract data from medical images and preprocess the extracted data, so as to organize them into a first data set, a second data set, and a third data set; The image segmentation and analysis module (2) is used to analyze the first data set, so as to generate an image segmentation qualified identification coefficient TFX, and analyze the image segmentation qualified identification coefficient TFX, so as to generate a first analysis result; The artifact analysis module (3) is used to analyze the second data set, so as to generate an artifact identification coefficient AIC, and analyze the artifact identification coefficient AIC, so as to generate a second analysis result; The 3D model analysis module (4) is used to integrally analyze the third data set and the artifact identification coefficient AIC, so as to generate a 3D modeling analysis coefficient MRC, and analyze the 3D modeling analysis coefficient MRC, so as to generate a third analysis result; The visualization operation module (5) is used to project the 3D model onto the visualization operation interface.

2. The medical image diagnosis assistance system based on animation design according to claim 1, characterized in that: The data acquisition module (1) includes a data extraction unit (11) and a data preprocessing unit (12); Among them, the data extraction unit (11) is used to perform multi-data acquisition on the image through acquisition software, including pixel resolution, slice thickness, contrast, gradient change rate, signal-to-noise ratio, gray-level co-occurrence matrix features, edge sharpness, tissue gray-level uniformity, scan time, scan energy, signal drift, stripe artifact intensity, metal artifact ratio, motion artifact amplitude, signal non-uniformity, scan angle error, voxel size, point cloud density, surface smoothness, topological connectivity, volume deviation, surface texture consistency, structure noise ratio, and model simplification error; The data preprocessing unit (12) preprocesses and dimensionlessizes the collected multiple data, and reorganizes them into a first data set, a second data set, and a third data set; The first data set includes pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix features HG, edge sharpness RD, and tissue gray-level uniformity ZH; The second data set includes scan time SS, scan energy SN, signal drift XP, stripe artifact intensity TW, metal artifact ratio JS, motion artifact amplitude YD, signal non-uniformity XB, and scan angle error WC; The third data set includes voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structure noise ratio ZB, and model simplification error MX.

3. The medical image diagnosis assistance system based on animation design according to claim 2, wherein: The image segmentation and analysis module (2) includes an image segmentation identification unit (21) and an image segmentation analysis unit (22); The image segmentation and identification unit (21) is used to perform integrated calculations on the first data set. By integrating the pixel resolution FB, slice thickness HD, contrast DB, gradient change rate TD, signal-to-noise ratio XZ, gray-level co-occurrence matrix feature HG, edge sharpness RD, and tissue gray-level uniformity ZH, the image segmentation qualified identification coefficient TFX is obtained. The image segmentation analysis unit (22) is used to compare the image segmentation qualified identification coefficient TFX with a preset image segmentation identification threshold Y1 to generate a first analysis result.

4. The medical image diagnosis assistance system based on animation design according to claim 3, wherein: The image segmentation and identification unit (21) calculates and obtains the image segmentation qualified identification coefficient TFX through the following formula: In the formula: FB is the pixel resolution, HD is the slice thickness, DB is the contrast, TD is the gradient change rate, XZ is the signal-to-noise ratio, HG is the gray-level co-occurrence matrix feature, RD is the edge sharpness, ZH is the tissue gray-level uniformity, and e is the base of the natural logarithm.

5. The medical image diagnosis assistance system based on animation design according to claim 4, wherein: The first analysis result generated by the image segmentation analysis unit (22) is specifically as follows: When AIC > Y1, it means that the current image segmentation area is abnormal and needs to be recalibrated. When AIC ≤ Y1, it means that the current image segmentation area is normal and the next step of processing is carried out.

6. The medical image diagnosis assistance system based on animation design according to claim 5, characterized in that: The artifact analysis module (3) includes an artifact identification unit (31) and an artifact analysis unit (32). The artifact identification unit (31) is used to calculate the second data set. By integrating the scan time SS, scan energy SN, signal drift XP, stripe artifact intensity TW, metal artifact ratio JS, motion artifact amplitude YD, signal inhomogeneity XB, and scan angle error WC, the artifact identification coefficient AIC is obtained. The artifact analysis unit (32) is used to compare the calculated artifact identification coefficient AIC with a preset artifact identification threshold Y2 to generate a second analysis result.

7. The medical image diagnosis assistance system based on animation design according to claim 6, characterized in that: The artifact identification unit (31) calculates and obtains the artifact identification coefficient AIC through the following formula: Where: SS is the scanning time, SN is the scanning energy, XP is the signal drift, TW is the intensity of stripe artifacts, JS is the ratio of metal artifacts, YD is the amplitude of motion artifacts, XB is the signal inhomogeneity, WC is the scanning angle error, a is the adjustment coefficient, and the value range of a is 0.52 to 2, log is the logarithmic function, and tan -1 is the arctangent function.

8. The medical image diagnosis assistance system based on animation design according to claim 7, characterized in that: The second analysis result generated by the artifact analysis unit (32) is specifically as follows: When AIC > Y2, it means that the current image has artifacts and needs to be denoised twice. When MRC ≤ Y2, it means that the current image has no artifacts and 3D modeling is carried out.

9. The medical image diagnosis assistance system based on animation design according to claim 8, wherein: The 3D model analysis module (4) includes a model identification and calculation unit (41) and a model analysis unit (42). The model identification and calculation unit (41) is used to perform integrated calculations on the third data set and the artifact identification coefficient AIC. By integrating the voxel size TS, point cloud density DY, surface smoothness QM, topological connectivity TP, volume deviation TC, surface texture consistency BM, structure noise ratio ZB, model simplification error MX, and artifact identification coefficient AIC, the 3D modeling analysis coefficient MRC is obtained. The model analysis unit ( ​ When MRC ≤ Y3, it means that the current 3D modeling does not meet the usage requirements and can be sent to the visualization terminal.

10. The medical image diagnosis assistance system based on animation design according to claim 9, characterized in that: The model identification calculation unit (41) calculates and obtains the 3D modeling analysis coefficient MRC through the following formula: Where: TS is the voxel size, DY is the point cloud density, QM is the surface smoothness, TP is the topological connectivity, TC is the volume deviation, BM is the surface texture consistency, ZB is the structure noise ratio, MX is the model simplification error, ln is the natural logarithm, and tan -1 is the arctangent function, and e is the base of the natural logarithm.

Images