{"title":"基于核磁共振成像脂肪抑制T2加权图像的多发性肌炎/皮肌炎肌肉水肿放射组学自动分级的价值。","authors":"Yumei Zhang, Yuefen Zou, Wenfeng Tan, C. Lv","doi":"10.1177/02841851241244507","DOIUrl":null,"url":null,"abstract":"BACKGROUND\nThe precise and objective assessment of thigh muscle edema is pivotal in diagnosing and monitoring the treatment of dermatomyositis (DM) and polymyositis (PM).\n\n\nPURPOSE\nRadiomic features are extracted from fat-suppressed (FS) T2-weighted (T2W) magnetic resonance imaging (MRI) of thigh muscles to enable automatic grading of muscle edema in cases of polymyositis and dermatomyositis.\n\n\nMATERIAL AND METHODS\nA total of 241 MR images were analyzed and classified into five levels using the Stramare criteria. The correlation between muscle edema grading and T2-mapping values was assessed using Spearman's correlation. The dataset was divided into a 7:3 ratio of training (168 samples) and testing (73 samples). Thigh muscle boundaries in FS T2W images were manually delineated with 3D-Slicer. Radiomics features were extracted using Python 3.7, applying Z-score normalization, Pearson correlation analysis, and recursive feature elimination for reduction. A Naive Bayes classifier was trained, and diagnostic performance was evaluated using receiver operating characteristic (ROC) curves and comparing sensitivity and specificity with senior doctors.\n\n\nRESULTS\nA total of 1198 radiomics parameters were extracted and reduced to 18 features for Naive Bayes modeling. In the testing set, the model achieved an area under the ROC curve of 0.97, sensitivity of 0.85, specificity of 0.98, and accuracy of 0.91. The Naive Bayes classifier demonstrated grading performance comparable to senior doctors. A significant correlation (r = 0.82, P <0.05) was observed between Stramare edema grading and T2-mapping values.\n\n\nCONCLUSION\nThe Naive Bayes model, utilizing radiomics features extracted from thigh FS T2W images, accurately assesses the severity of muscle edema in cases of PM/DM.","PeriodicalId":7143,"journal":{"name":"Acta radiologica","volume":null,"pages":null},"PeriodicalIF":1.1000,"publicationDate":"2024-04-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Value of radiomics-based automatic grading of muscle edema in polymyositis/dermatomyositis based on MRI fat-suppressed T2-weighted images.\",\"authors\":\"Yumei Zhang, Yuefen Zou, Wenfeng Tan, C. Lv\",\"doi\":\"10.1177/02841851241244507\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"BACKGROUND\\nThe precise and objective assessment of thigh muscle edema is pivotal in diagnosing and monitoring the treatment of dermatomyositis (DM) and polymyositis (PM).\\n\\n\\nPURPOSE\\nRadiomic features are extracted from fat-suppressed (FS) T2-weighted (T2W) magnetic resonance imaging (MRI) of thigh muscles to enable automatic grading of muscle edema in cases of polymyositis and dermatomyositis.\\n\\n\\nMATERIAL AND METHODS\\nA total of 241 MR images were analyzed and classified into five levels using the Stramare criteria. The correlation between muscle edema grading and T2-mapping values was assessed using Spearman's correlation. The dataset was divided into a 7:3 ratio of training (168 samples) and testing (73 samples). Thigh muscle boundaries in FS T2W images were manually delineated with 3D-Slicer. Radiomics features were extracted using Python 3.7, applying Z-score normalization, Pearson correlation analysis, and recursive feature elimination for reduction. A Naive Bayes classifier was trained, and diagnostic performance was evaluated using receiver operating characteristic (ROC) curves and comparing sensitivity and specificity with senior doctors.\\n\\n\\nRESULTS\\nA total of 1198 radiomics parameters were extracted and reduced to 18 features for Naive Bayes modeling. In the testing set, the model achieved an area under the ROC curve of 0.97, sensitivity of 0.85, specificity of 0.98, and accuracy of 0.91. The Naive Bayes classifier demonstrated grading performance comparable to senior doctors. A significant correlation (r = 0.82, P <0.05) was observed between Stramare edema grading and T2-mapping values.\\n\\n\\nCONCLUSION\\nThe Naive Bayes model, utilizing radiomics features extracted from thigh FS T2W images, accurately assesses the severity of muscle edema in cases of PM/DM.\",\"PeriodicalId\":7143,\"journal\":{\"name\":\"Acta radiologica\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":1.1000,\"publicationDate\":\"2024-04-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Acta radiologica\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1177/02841851241244507\",\"RegionNum\":4,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Acta radiologica","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1177/02841851241244507","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}
Value of radiomics-based automatic grading of muscle edema in polymyositis/dermatomyositis based on MRI fat-suppressed T2-weighted images.
BACKGROUND
The precise and objective assessment of thigh muscle edema is pivotal in diagnosing and monitoring the treatment of dermatomyositis (DM) and polymyositis (PM).
PURPOSE
Radiomic features are extracted from fat-suppressed (FS) T2-weighted (T2W) magnetic resonance imaging (MRI) of thigh muscles to enable automatic grading of muscle edema in cases of polymyositis and dermatomyositis.
MATERIAL AND METHODS
A total of 241 MR images were analyzed and classified into five levels using the Stramare criteria. The correlation between muscle edema grading and T2-mapping values was assessed using Spearman's correlation. The dataset was divided into a 7:3 ratio of training (168 samples) and testing (73 samples). Thigh muscle boundaries in FS T2W images were manually delineated with 3D-Slicer. Radiomics features were extracted using Python 3.7, applying Z-score normalization, Pearson correlation analysis, and recursive feature elimination for reduction. A Naive Bayes classifier was trained, and diagnostic performance was evaluated using receiver operating characteristic (ROC) curves and comparing sensitivity and specificity with senior doctors.
RESULTS
A total of 1198 radiomics parameters were extracted and reduced to 18 features for Naive Bayes modeling. In the testing set, the model achieved an area under the ROC curve of 0.97, sensitivity of 0.85, specificity of 0.98, and accuracy of 0.91. The Naive Bayes classifier demonstrated grading performance comparable to senior doctors. A significant correlation (r = 0.82, P <0.05) was observed between Stramare edema grading and T2-mapping values.
CONCLUSION
The Naive Bayes model, utilizing radiomics features extracted from thigh FS T2W images, accurately assesses the severity of muscle edema in cases of PM/DM.
期刊介绍:
Acta Radiologica publishes articles on all aspects of radiology, from clinical radiology to experimental work. It is known for articles based on experimental work and contrast media research, giving priority to scientific original papers. The distinguished international editorial board also invite review articles, short communications and technical and instrumental notes.