Fig. 2From: Automatic segmentation of large-scale CT image datasets for detailed body composition analysisSection A: Illustration of IPAT mask generation, A (Abdomen CT slice), B (VAT) model output, C (RPAT) model output, D (IPAT). Section B: Illustration of DSAT and SSAT mask generation. A (Abdomen CT Slice), B (SAT) model output, C (Raw mask for DSAT), D (SSAT and DSAT). Section C: Illustration of segmentation of cortical bone and bone marrow. A (Thigh CT slice), B (Cortical bone), C (Bone marrow), D (segmented output overlayed on CT image)Back to article page