MOOSE (Multi-organ objective segmentation) a data-centric AI solution that generates multilabel organ segmentations to facilitate systemic TB whole-person research.The pipeline is based on nn-UNet and has the capability to segment 120 unique tissue classes from a whole-body 18F-FDG PET/CT image.
3d-segmentation brain-segmentation pet-ct ct-segmentation nnunet organ-segmentation tb-pet bone-segmentation muscle-segmentation fat-segmentation psoas-segmentation total-body-segmentation
-
Updated
Oct 31, 2026 - Python