Computed Tomography-Based Artificial Intelligence Analysis of Myosteatosis and Its Effect on Survival in Fournier Gangrene

PURPOSE: Fournier gangrene (FG) is a life-threatening necrotizing infection with a high mortality rate. Conventional prognostic indices, including the Fournier Gangrene Severity Index (FGSI), depended on static parameters and may not adequately capture baseline metabolic vulnerability. Myosteatosis, which is the pathological infiltration of fat within skeletal muscle, has been associated with poor outcomes in cancer and critical illness. This study evaluated the prognostic significance of myosteatosis in FG using artificial intelligence (AI)-based computed tomography (CT) analysis. MATERIALS AND METHODS: We retrospectively analyzed 40 male patients with FG, whose diagnosis was confirmed based on clinical, radiological, and surgical findings. Preoperative CT scans at the L3 level were assessed using AI-based software to quantify skeletal muscle quality. Quantitative parameters were automatically extracted to evaluate muscle quality. Patients were classified into myosteatosis or non-myosteatosis groups according to established cut-off values. Clinical characteristics, FGSI, septic shock, and in-hospital mortality were compared between groups. Predictors of in-hospital mortality were evaluated using multivariate logistic regression. RESULTS: 56.5%±15.2%, p9, and cardiovascular disease as independent predictors of in-hospital mortality. CONCLUSIONS: AI-based CT analysis of myosteatosis is a reproducible method for assessing muscle quality in the FG. Myosteatosis was independently associated with septic shock and in-hospital mortality beyond body mass index and FGSI, suggesting its potential as a novel imaging biomarker for early risk stratification and individualized treatment planning.