Intelligent diagnosis system and method for Crohn disease muscular layer change

Abstract

The invention discloses an intelligent diagnosis system and method for Crohn disease muscular layer change, and belongs to the technical field of medical image analysis and artificial intelligence, and the system comprises an image collection and processing module which is used for receiving an MRI image of a patient, carrying out the de-noising and enhanced comparison, providing high-quality image data for the subsequent scoring analysis, and carrying out the calculation of the image data based on a preset algorithm, automatically sketching a target area; the image feature extraction module is used for extracting 102 radiomics features from the T1 enhanced image by using a Pyradiomics tool, and combining DCE-MRI functional parameters and conventional magnetic resonance parameters; the clinical data integration module is used for integrating clinical indexes of patients and constructing a multi-dimensional feature pool; and the score calculation module is used for calculating the extracted features. According to the intelligent diagnosis system and method for the Crohn disease muscular layer change, scientific research development can be promoted, an innovative muscular layer change evaluation scheme provides reliable data support for further scientific research, and the research progress in the field can be promoted.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical image analysis and artificial intelligence, and in particular relates to an intelligent diagnosis system for muscular layer changes in Crohn's disease, and also relates to an intelligent diagnosis method for muscular layer changes in Crohn's disease. Background Art

[0002] Crohn's disease is a chronic inflammatory bowel disease characterized by transmural intestinal inflammation and often complicated by intestinal stenosis. More and more studies have shown that in the pathological mechanism of intestinal stenosis, smooth muscle hyperplasia / hypertrophy is the main factor leading to fibrotic stricture lesions.

[0003] In the prior art, pathological examination after intestinal resection surgery is still the gold standard for histopathological diagnosis of Crohn's disease. However, pathological diagnosis can only be carried out after surgical treatment and cannot be used as an indicator for dynamically evaluating the histopathological changes of Crohn's disease lesions.

[0004] Therefore, it is necessary to develop an intelligent diagnosis system and method for muscular layer changes in Crohn's disease to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent diagnosis system and method for muscular layer changes in Crohn's disease to solve the problems in the background art, such as the diagnosis of histopathological changes in Crohn's disease relying on postoperative pathological examination and the lack of non-invasive standardized tools.

[0006] To achieve the above purpose, the present invention provides the following technical solution: An intelligent diagnosis system for muscular layer changes in Crohn's disease, comprising:

[0007] An image acquisition and processing module: Receiving the patient's MRI image, performing denoising and contrast enhancement, providing high-quality image data for subsequent scoring analysis, and automatically delineating the target area based on a preset algorithm;

[0008] An image feature extraction module: Extracting 102 radiomics features from the T1-weighted enhanced image by using Pyradiomics tools, and combining DCE-MRI functional parameters and conventional magnetic resonance parameters;

[0009] A clinical data integration module: Integrating the patient's clinical indicators to construct a multi-dimensional feature pool;

[0010] A scoring calculation module: According to the extracted features, giving different weights to different parameters according to the muscular layer change evaluation model and nomogram, and automatically calculating and generating a scoring result;

[0011] A result display and report generation module: Displaying the scoring result to the doctor in an intuitive manner and automatically generating a detailed scoring report.

[0012] As a preferred embodiment, the patient's MRI images include T1-weighted enhancement, DWI, and DCE-MRI sequences. The target region includes the entire diseased intestinal wall, excluding the gas in the intestinal lumen and the adjacent tissues outside the intestinal wall.

[0013] As a preferred embodiment, the radiomics features include gray-level co-occurrence matrix (GLCM) and gray-level run length matrix (GLRLM). The DCE-MRI functional parameters include Ktrans, Ve, and Kep. The conventional magnetic resonance parameters include T2ratio and ADC.

[0014] As a preferred embodiment, the patient's clinical indicators include CRP, CDAI, disease course, BMI, and complications. The scoring report includes scoring details and image comparison, facilitating the doctor's disease analysis and treatment decision-making.

[0015] An intelligent diagnosis method for muscular layer changes in Crohn's disease, used to evaluate the muscular layer changes in Crohn's disease and give corresponding scores according to the model, includes the following steps:

[0016] Step 1: Import the patient's MRI images into the tool.

[0017] Step 2: The tool automatically performs image preprocessing and feature extraction.

[0018] Step 3: The tool extracts the patient's clinical indicators.

[0019] Step 4: According to the extracted features and clinical indicators, automatically calculate and generate the scoring result.

[0020] Step 5: Display the scoring result and generate a detailed scoring report for the doctor's reference.

[0021] As a preferred embodiment, the patient's clinical indicators include age, CRP, CDAI, disease course, and BMI.

[0022] Compared with the prior art, the technical effects and advantages of the present invention:

[0023] The intelligent diagnosis system and method for muscular layer changes in Crohn's disease can improve the scoring efficiency. The automated scoring process greatly shortens the scoring time and improves the work efficiency.

[0024] The intelligent diagnosis system and method for muscular layer changes in Crohn's disease can enhance the scoring accuracy, reduce the interference of human factors, and improve the objectivity and accuracy of the scoring.

[0025] The intelligent diagnosis system and method for muscular layer changes in Crohn's disease can assist in clinical decision-making, provide a more accurate and comprehensive basis for the doctor to evaluate the disease condition, and help formulate a more reasonable treatment plan.

[0026] The intelligent diagnosis system and method for muscular layer changes in Crohn's disease can promote scientific research development. The innovative evaluation scheme for muscular layer changes provides reliable data support for further scientific research and helps to promote the research progress in this field. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a framework diagram of the present invention.

[0028] Figure 2 This is an example of a nomogram, quantifying the contribution weights of radiomics features, DCE-MRI parameters, and clinical indicators. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] In the following description, numerous specific details are given to provide a more thorough understanding of the present invention. However, it is obvious to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, some well-known technical features are not described to avoid confusion with the present invention.

[0030] Please refer to Figures 1 to 2 , an intelligent diagnosis system for muscular layer changes in Crohn's disease. In this embodiment, it includes:

[0031] Image acquisition and processing module: Receives the patient's MRI images, performs denoising and contrast enhancement to provide high-quality image data for subsequent scoring analysis, and automatically outlines the target area based on a preset algorithm;

[0032] Image feature extraction module: Extracts 102 radiomics features from the T1-weighted enhanced image using Pyradiomics tools, and combines DCE-MRI functional parameters and conventional magnetic resonance parameters;

[0033] Clinical data integration module: Integrates the patient's clinical indicators to construct a multi-dimensional feature pool;

[0034] Scoring calculation module: According to the extracted features, gives different weights to different parameters according to the muscular layer change evaluation model and nomogram, and automatically calculates and generates a scoring result;

[0035] Result display and report generation module: Displays the scoring result to the doctor in an intuitive manner and automatically generates a detailed scoring report.

[0036] The patient's MRI images include T1-weighted enhanced, DWI, and DCE-MRI sequences. The target area includes the entire intestinal wall of the lesion, excluding the gas in the intestinal lumen and the adjacent tissues outside the intestinal wall.

[0037] The radiomics features include Gray-Level Co-Occurrence Matrix (GLCM) and Gray-Level Run-Length Matrix (GLRLM). The DCE-MRI functional parameters include Ktrans, Ve, and Kep. The conventional magnetic resonance parameters include T2ratio and ADC.

[0038] The patient clinical indicators include CRP, CDAI, disease course, BMI, and complications. The scoring report includes scoring details and image comparison, facilitating doctors' disease analysis and treatment decision-making.

[0039] An intelligent diagnosis method for the muscular layer changes in Crohn's disease, used to evaluate the muscular layer changes in Crohn's disease and give corresponding scores according to the model, includes the following steps:

[0040] Step 1: Import the patient's MRI images into the tool;

[0041] Step 2: The tool automatically performs image preprocessing and feature extraction;

[0042] Step 3: The tool extracts the patient's clinical indicators;

[0043] Step 4: According to the extracted features and clinical indicators, automatically calculate and generate a scoring result;

[0044] Step 5: Display the scoring result and generate a detailed scoring report for doctors' reference.

[0045] The patient clinical indicators include age, CRP, CDAI, disease course, and BMI.

[0046] The detailed implementation steps of the scoring tool:

[0047] S1, System initialization and image import: Start the intelligent scoring system for the muscular layer changes in Crohn's disease, load the preset scoring protocol and algorithm model. At the same time, import the patient's abdominal DCE-MRI images into the system; these images should contain clear intestinal tract structures and have undergone necessary preprocessing such as denoising and contrast enhancement;

[0048] S2, Image preprocessing and segmentation: The system uses deep learning algorithms to further preprocess the imported images, including image registration, artifact removal, and identification and automatic segmentation of the Crohn's disease lesion area. This step aims to accurately extract the Crohn's disease lesion area from the complex background, laying a foundation for subsequent analysis;

[0049] S3, Feature extraction and quantification: Use the Pyradiomics tool to extract 102 radiomics features from the T1-weighted enhanced images, including Gray-Level Co-Occurrence Matrix (GLCM), Gray-Level Run-Length Matrix (GLRLM), etc., and extract the DCE-MRI functional parameters and conventional magnetic resonance parameters. The system will quantify these features;

[0050] S4, Clinical data integration: The tool extracts the patient's clinical indicators from the patient case system and constructs a multi-dimensional feature pool;

[0051] S5, Calculation of muscular layer change score: Based on the extracted feature dataset, the system uses a pre-trained scoring model and combines it with a nomogram to automatically calculate the muscular layer change score;

[0052] S6, Display of scoring results and report generation: The system presents the automatically generated muscular layer change scoring results to the user in an intuitive way, such as through charts, color coding, etc. At the same time, the system will also generate a detailed report containing the scoring results, image analysis details, and diagnostic suggestions, facilitating doctors to quickly understand the pathological conditions of the muscular layer changes in the Crohn's disease lesions of the patients;

[0053] S7, Data storage and remote access: The scoring results and the generated reports will be securely stored in the system's database for subsequent reference and scientific research analysis. In addition, the system also supports the remote access function, and doctors can view the patients' scoring results and reports anytime and anywhere through the network.

[0054] To further improve the system's functions, the following are the detailed supplements to its workflow:

[0055] 1. In step S1, the scoring protocol and algorithm model loaded by the system should be based on the latest research results of Crohn's disease, and the model parameters can be adjusted subsequently according to research progress and follow-up;

[0056] 2. In step S2, the accuracy of image segmentation is crucial for subsequent analysis. The system should adopt advanced deep learning algorithms, such as U-Net, MaskR-CNN, etc., to improve the segmentation accuracy;

[0057] 3. In step S3, the process of feature extraction and quantization is unified, avoiding errors caused by manual operations;

[0058] 4. In step S4, the scoring system automatically extracts the clinical data in the case, calculates, reports missing values, and provides a window for manual input;

[0059] 5. In step S5, the scoring model should be trained, verified with a large amount of data, and continuously updated to ensure the accuracy of scoring. At the same time, the system should also provide the function of manually adjusting the score to meet the scoring requirements in special cases;

[0060] 6. In step S6, the report generated by the system should contain rich image analysis details and diagnostic suggestions to assist doctors in making more accurate diagnosis and treatment decisions;

[0061] 7. In step S7, the system shall ensure the security and privacy of data by adopting encryption storage and transmission technologies to prevent data leakage.

[0062] In addition, the system shall also have the following characteristics:

[0063] High degree of automation: The system can automatically complete the whole process of image preprocessing, feature extraction, score calculation and report generation, reducing manual intervention and improving work efficiency;

[0064] High accuracy: The system uses deep learning algorithms for image analysis and score calculation, can evaluate the pathological changes of Crohn's disease lesions, and give reliable score results;

[0065] Strong scalability: The system supports the update and optimization of algorithm models to adapt to new research results and clinical needs;

[0066] Strong user-friendliness: The system interface is simple and clear, the operation is convenient, and it is easy for doctors to get started.

[0067] This software not only shows extremely strong generalization in design. By supporting multiple image formats (such as DICOM, etc.) and adapting to magnetic resonance devices of different brands and models, it ensures seamless docking in various medical environments. It helps doctors complete the score and evaluation of the muscular layer changes of Crohn's disease more efficiently and accurately. This will not only greatly improve the accuracy of medical diagnosis and the personalization of treatment plans, but also is expected to promote the standardization and sharing of medical data, and further promote the scientific research progress and clinical practice development in the field of Crohn's disease.

[0068] Detailed definition of the input end and patent protection scope

[0069] For the intelligent diagnosis system of Crohn's disease muscular layer changes based on the fusion of multi-modal MRI radiomics and clinical data described in the present invention, in the design of the input end, wide compatibility and security are fully considered, and the specific definition is as follows:

[0070] 1. Supported file formats: This tool strictly follows the international standards in the field of medical imaging, mainly supports MRI image files in DICOM (version 3.0 and above) format. The DICOM file shall contain complete metadata, including but not limited to key information such as patient name (PatientName), examination date (StudyDate), sequence description (SequenceDescription), etc., to ensure the integrity and traceability of image data.

[0071] 2. Input method:

[0072] Local File Import: Users can select DICOM files stored in the local file system through the Graphical User Interface (GUI). The files should be located in a preset folder, and the file names need to follow the naming rule of "Patient ID_Examination Date_Timestamp.dcm" to ensure the orderliness of file management and the efficiency of data retrieval.

[0073] Network Transmission: This tool also provides an HTTP REST API interface, which supports remote systems to send image data through POST requests. The request body should be in JSON format, containing the Base64-encoded content of the DICOM image file and necessary metadata fields such as patient ID, examination type, etc. The API interface uses the HTTPS protocol to ensure the security and integrity of the data transmission process.

[0074] 3. User Interface Design: The GUI interface is designed to be intuitive and easy to use, including elements such as clear file import buttons, progress indicators, error prompt messages, etc. Users only need to click the import button and select or enter the file path to automatically complete the loading and preprocessing of the image data.

[0075] 4. Claims: The claims of the present invention are clearly defined. Any unauthorized modification of the input format (including but not limited to DICOM version, metadata requirements), file naming rule, API interface definition (including request method, parameter type, return value format), or GUI interface layout is regarded as an infringement of the patent right of the present invention. The protection scope of this patent is not limited to the above specific implementation manners, but also includes any substantially equivalent input means and technical solutions.

[0076] 5. Examples and Illustrations: For the convenience of understanding and implementation, the patent application document is accompanied by screenshots of DICOM file samples, GUI interface layout diagrams, and API call example codes to intuitively show the specific requirements and operation processes at the input end.

[0077] In summary, through the detailed and specific definition of the input end, the present invention aims to ensure the wide applicability of the automatic scoring tool in different medical environments and the effectiveness of patent protection.

[0078] Crohn's disease

[0079] Crohn’s Disease (CD) is a chronic inflammatory bowel disease characterized by transmural intestinal inflammation. Its pathological changes not only involve the mucosal layer but also often accompany significant abnormalities in the intestinal muscular layer. Muscular layer changes are important driving factors for intestinal stricture and fibrosis in Crohn’s disease. Its research involves multiple aspects such as etiology, clinical manifestations, diagnostic methods, treatment strategies, and evaluation tools. The following is an extended introduction to the muscular layer changes in Crohn’s disease:

[0080] I. Etiology

[0081] The specific mechanism of muscular layer changes in Crohn's disease has not been fully elucidated, but the interaction of genetic, immune, and environmental factors is considered crucial. The NOD2 / CARD15 gene mutation is closely related to the susceptibility of Crohn's disease and may affect muscular layer remodeling by regulating intestinal immune responses and myofibroblast differentiation. Chronic inflammation leads to the over-secretion of profibrotic cytokines (such as TGF-β, IL-13), driving the hyperplasia and hypertrophy of smooth muscle cells. Intestinal microbiota dysbiosis may exacerbate the inflammatory response and the process of muscular layer fibrosis by activating the innate immune system.

[0082] II. Clinical Manifestations

[0083] Typical manifestations of muscular layer changes in Crohn's disease include: thickening of the intestinal wall and stenosis of the lumen caused by smooth muscle hyperplasia / hypertrophy, leading to recurrent abdominal pain, intestinal obstruction, and abdominal distension. Under the combined action of inflammation and fibrosis, the elasticity of the intestine is lost, forming an irreversible "rigid intestinal segment". Muscular layer inflammation can penetrate the intestinal wall, forming fistulas or abscesses, increasing the risk of perforation.

[0084] III. Diagnostic Methods

[0085] The diagnosis of muscular layer changes requires the combination of imaging, endoscopic, and pathological techniques: 1) High-resolution T2-weighted sequences and dynamic contrast-enhanced scans of MRE can clearly show intestinal wall thickening, muscular layer edema, and fibrosis; 2) Ultrasound elastography can evaluate the degree of fibrosis by measuring the hardness of the intestinal wall; 3) Endoscopic examination can show intestinal lumen stenosis and the formation of mucosal bridges, but it is difficult to directly evaluate muscular layer lesions; 4) Pathological Masson trichrome staining of surgical specimens can distinguish smooth muscle hyperplasia from collagen deposition, but it relies on invasive biopsies.

[0086] IV. Treatment Strategies

[0087] The treatment for muscular layer changes aims to relieve stenosis and delay the process of fibrosis: Biological agents, anti-TNF-α drugs (such as infliximab), can inhibit inflammation and indirectly slow down muscular layer remodeling. New drugs targeting TGF-β or integrins (such as STX-100) are in clinical trials. Endoscopic balloon dilation is applicable to local stenosis, but the recurrence rate is high. Surgical segmental intestinal resection is used for severe stenosis or obstruction, but it may accelerate the progression of lesions in the remaining intestine.

[0088] V. Disease Assessment Tools

[0089] The assessment of the activity of Crohn's disease requires multi-dimensional integration, and the core indicators include: clinical symptom scores (CDAI / HBI), inflammatory markers (CRP, fecal calprotectin), imaging (MRE), and endoscopy (SES-CD), etc.

[0090] VI. Research Progress

[0091] With the continuous deepening of medical research, the understanding of Crohn's disease has gradually become clearer from being vague, and the treatment methods have also been continuously innovated. For example, the emergence of biological agent treatment has provided new treatment options for Crohn's disease patients and significantly improved the prognosis of patients. At the same time, the in-depth study of the muscular layer changes in Crohn's disease has also made it possible to develop new treatment targets. In addition, with the development of medical technology, the development and application of automated scoring tools will further improve the accuracy and efficiency of the assessment of Crohn's disease.

[0092] In summary, the research on Crohn's disease involves multiple aspects, including etiology, clinical manifestations, diagnostic methods, treatment strategies, and disease assessment tools, etc. With the continuous progress of medical research and the development of technology, it is believed that more new treatment methods and assessment tools will emerge in the future, bringing better treatment effects and quality of life to Crohn's disease patients.

[0093] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent diagnosis system for muscular layer changes in Crohn's disease, characterized in that: Including: Image acquisition and processing module: Receives the patient's MRI images, performs denoising and contrast enhancement, provides high-quality image data for subsequent scoring analysis, and automatically delineates the target area based on a preset algorithm; Image feature extraction module: Extracts 102 radiomics features from the T1-weighted enhanced images using Pyradiomics tools, and combines DCE-MRI functional parameters and conventional magnetic resonance parameters; Clinical data integration module: Integrates the patient's clinical indicators to construct a multi-dimensional feature pool; Scoring calculation module: According to the extracted features, assigns different weights to different parameters according to the muscularis propria change assessment model and nomogram, and automatically calculates and generates a scoring result; Result display and report generation module: Displays the scoring result to the doctor in an intuitive manner and automatically generates a detailed scoring report.

2. The intelligent diagnosis system for muscular layer changes of Crohn's disease according to claim 1, characterized in that: The patient's MRI images include T1-weighted enhanced, DWI, and DCE-MRI sequences. The target area includes the entire lesion intestinal wall, excluding the gas in the intestinal lumen and the adjacent tissues outside the intestinal wall.

3. The intelligent diagnosis system for muscular layer changes of Crohn's disease according to claim 1, wherein: The radiomics features include gray-level co-occurrence matrix (GLCM) and gray-level run length matrix (GLRLM). The DCE-MRI functional parameters include Ktrans, Ve, and Kep. The conventional magnetic resonance parameters include T2ratio and ADC.

4. The intelligent diagnosis system for muscular layer changes in Crohn's disease according to claim 1, characterized in that: The patient's clinical indicators include CRP, CDAI, disease course, BMI, and complications. The scoring report includes scoring details and image comparison, facilitating the doctor's condition analysis and treatment decision-making.

5. An intelligent diagnosis method for muscular layer changes in Crohn's disease, which is used to operate and implement any one of the intelligent diagnosis systems for muscular layer changes in Crohn's disease described in claims 1-4, and is used to evaluate the muscular layer changes in Crohn's disease and give corresponding scores according to the model, characterized in that: Including the following steps: Step 1: Import the patient's MRI images into the tool; Step 2: The tool automatically performs image preprocessing and feature extraction; Step 3: The tool extracts the patient's clinical indicators; Step 4: According to the extracted features and clinical indicators, automatically calculates and generates a scoring result; Step 5: Displays the scoring result and generates a detailed scoring report for the doctor's reference.

6. The intelligent diagnosis method for muscular layer changes of Crohn's disease according to claim 5, characterized in that: The patient's clinical indicators include age, CRP, CDAI, disease course, and BMI.

Images