Machine learning models based on preoperative contrast-enhanced CT radiomics for differential diagnosis of xanthogranulomatous cholecystitis and thick-walled gallbladder carcinoma: A retrospective study

Background Xanthogranulomatous cholecystitis (XGC) is a rare form of gallbladder chronic inflammatory disease that mimics thick-walled gallbladder carcinoma (GBC), thereby affecting treatment decisions. We aimed to analyze the radiological characteristics of XGC and thick-walled GBC to establish a diagnostic prediction model for differential diagnosis and treatment decisions. Methods A total of 129 patients with gallbladder disease who underwent upper abdominal enhanced CT and curative-intent resection in a high-volume medical center from January 2011 to December 2020 were retrospectively analyzed. Results There were significant differences in the CT imaging features of gallbladder wall thickening pattern, mucosal lines, intramural nodules, gallstones, abnormal enhancement of adjacent liver parenchyma, peripheral enlargement lymph nodes between XGC and thick-walled GBC ( P