Abstract:Purpose: To perform a radiomics analysis with comparisons of multidomain features and a variety of feature selection strategies and classifiers, with the goal of evaluating the value of quantified radiomics method for noninvasively differentiating autoimmune pancreatitis (AIP) from pancreatic ductal adenocarcinoma (PDAC) in 18 F-fluorodeoxglucose positron emission tomography/computed tomography (18 F FDG PET/CT) images. Methods: We extracted 251 expert-designed features from 2D and 3D PET/CT images of 111 patients, and recombined these features into five feature sets according to their modalities and dimensions. Among the five feature sets, the optimal one was found leveraging four feature selection strategies and four machine learning classifiers based on the area under the receiver operating characteristic curve (AUC). The feature selection strategies include spearman's rank correlation coefficient, minimum redundancy maximum relevance, support vector machine recursive feature elimination (SVM-RFE), and no feature selection, while the classifiers are random forest, adaptive boosting, support vector machine (SVM) with the Gaussian radial basis function, and SVM with the linear kernel function respectively. Based on the optimal feature set, these feature selection strategies and classifiers were comparatively studied to achieve the best differentiation. Finally, the quantified radiomics prediction model was developed based on the best combination of the feature selection strategy and classifier, and it was compared with two clinical factors based prediction models, and human doctors using nested cross-validation in terms of AUC, accuracy, sensitivity, and specificity. Results: Comparison experiments demonstrated that CT features and three-dimensional (3D) features performed better than positron emission tomography (PET) features and three-dimensional (2D) features respectively, and multidomain features were superior to single domain features. In addition, the combination of SVM-RFE feature selection strategy and Linear SVM classifier had the highest diagnostic performance (i.e., AUC = 0.93 ± 0.01, ACC = 0.85 ± 0.02, SEN = 0.86 ± 0.03, SPE = 0.84 ± 0.03). The quantified radiomics model developed is significantly superior to both human doctors and clinical factors based prediction models in terms of accuracy and specificity. Conclusions: Our preliminary results confirmed that the quantified radiomics method could aid the noninvasive differentiation of AIP and PDAC in 18 F FDG PET/CT images and the integration of multidomain features is beneficial for the differentiation.

Radiological biomarkers reflecting visceral fat distribution help distinguish inflammatory bowel disease subtypes: a multicenter cross-sectional study

Volumetric visceral fat machine learning phenotype on CT for differential diagnosis of inflammatory bowel disease

A Novel Computed Tomography Enterography Radiomics Combining Intestinal and Creeping Fat Features Could Predict Surgery Risk in Patients with Crohn’s Disease

AI-based Fingerprint Index of Visceral Adipose Tissue for the Prediction of Bowel Damage in Patients with Crohn's Disease

CT-based radiomics signature of visceral adipose tissue and bowel lesions for identifying patients with Crohn's disease resistant to infliximab

P1035 CT-based radiomics signature of visceral adipose tissue and bowel lesions for identifying patients with Crohn’s disease resistant to infliximab

Enhancing radiologist's detection: an imaging-based grading system for differentiating Crohn's disease from ulcerative colitis

Deciphering CT texture features of human visceral fat to evaluate metabolic disorders and surgery-induced weight loss effects

Assessing Visceral Obesity and Abdominal Adipose Tissue Distribution in Healthy Populations Based on Computed Tomography: A Large Multicenter Cross-Sectional Study.

Comparison of the Value of Adipose Tissues in Abdomen and Lumbar Vertebra for Predicting Disease Activity in Crohn's Disease: A Preliminary Study Based on CSE-MRI

A Mesenteric Fat-Derived Radiomic Model to Identify Colonic Fibrosis and Predict Treatment Response to Biologics in Chronic Ulcerative Colitis

Radiomics analysis for the differentiation of autoimmune pancreatitis and pancreatic ductal adenocarcinoma in 18 F-FDG PET/CT

Enhancing the Differentiation between Intestinal Beh?et's Disease and Crohn's Disease through Quantitative Computed Tomography Analysis

A radiomics nomogram based on MSCT and clinical factors can stratify fibrosis in inflammatory bowel disease

Towards visceral fat estimation at population scale: correlation of visceral adipose tissue assessment using three-dimensional cross-sectional imaging with BIA, DXA, and single-slice CT

Computed tomography-based multiple body composition parameters predict outcomes in Crohn’s disease

A novel computed tomography enterography radiomics combining intestinal and creeping fat features could predict surgery risk in patients with Crohn's disease

P824 Association between visceral adipose tissue and disease duration, clinical activity, and response to various biological agents in inflammatory bowel disease

Machine learning-based radiomic features of perivascular adipose tissue in coronary computed tomography angiography predicting inflammation status around atherosclerotic plaque: a retrospective cohort study

Creeping fat in patients with ileo-colonic Crohn's disease correlates with disease activity and severity of inflammation: A preliminary study using energy spectral computed tomography

Assessment of whole-body and regional body fat using abdominal quantitative computed tomography in Chinese women and men