Automated visceral adipose tissue (VAT) quantification from full-body MRI scans.
ScatFat analyzes whole-body MRI DICOM series to segment and measure visceral and subcutaneous fat compartments. Visceral fat (VAT) is a stronger predictor of cardiometabolic disease than BMI or total body fat — and direct MRI quantification gets you a measurement you can actually trust.
The pipeline auto-detects Dixon in-phase / opposed-phase series, runs TotalSegmentator (MRI tissue task) for segmentation, classifies the masks into VAT and SAT, computes a Dixon fat-signal-fraction diagnostic, and outputs volumes (cm³) and mass (kg) using adipose-tissue density of 0.923 g/cm³.
When vertebral landmarks are present, ScatFat also produces a DXA-comparable VAT proxy cropped to roughly T12 → L4 — so you can compare against DXA-based estimates while knowing where the methodologies diverge.