Deep learning-derived CT body composition enhances survival risk stratification beyond the TNM system in locally advanced gastric cancer: a multi-modality cohort study.

Key Points

• To evaluate the effectiveness of deep learning-derived CT body composition metrics in improving survival risk stratification for patients with locally advanced gastric cancer.
• Analyzed CT body composition metrics derived from deep learning techniques
• Compared prognostic capabilities of VAT/SAT ratio to traditional TNM staging
• Evaluated the relationship between body composition and immunometabolic phenotypes
• VAT/SAT ratio significantly improved risk stratification compared to TNM staging
• Identified mitochondrial dysfunction and immune suppression as important factors associated with the VAT/SAT ratio
• Recommends VAT/SAT as a noninvasive biomarker for guiding personalized therapy in gastric cancer management