VIBESegmentator: Full Body MRI Segmentation for the NAKO and UK Biobank

TL;DR

VIBESegmentator is a comprehensive deep learning model for full-body MRI and CT segmentation, covering a wide range of anatomical structures with high accuracy, and is publicly available for research use.

Contribution

The paper introduces a novel, publicly accessible deep learning model that segments nearly all torso voxels in MRI and CT images, extending previous models with broader coverage and improved accuracy.

Findings

Achieved an average Dice score of 0.90 on internal MRI test set.

Tied with the best model on Amos dataset with Dice of 0.81.

Segmented 71 structures in MRI and 72 in CT images, including organs, muscles, vessels, bones, and body composition.

Abstract

Objectives: To present a publicly available deep learning-based torso segmentation model that provides comprehensive voxel-wise coverage, including delineations that extend to the boundaries of anatomical compartments. Materials and Methods: We extracted preliminary segmentations from TotalSegmentator, spine, and body composition models for Magnetic Resonance Tomography (MR) images, then improved them iteratively and retrained an nnUNet model. Using a random retrospective subset of German National Cohort (NAKO), UK Biobank, internal MR and Computed Tomography (CT) data (Training: 2897 series from 626 subjects, 290 female; mean age 53+-16; 3-fold-cross validation (20% hold-out). Internal testing 36 series from 12 subjects, 6 male; mean age 60+-11), we segmented 71 structures in torso MR and 72 in CT images: 20 organs, 10 muscles, 19 vessels, 16 bones, ribs in CT, intervertebral discs, spinal cord, spinal canal and body composition (subcutaneous fat, unclassified muscles and visceral fat). For external validation, we used existing automatic organ segmentations, independent ground truth segmentations on gradient echo images, and the Amos data. We used non-parametric bootstrapping for confidence intervals and Wilcoxon rank-sum test for computing statistical significance. Results: We achieved an average Dice score of 0.90+-0.06 on our internal gradient echo test set, which included 71 semantic segmentation labels. Our model ties with the best model on Amos with a Dice of 0,81+-0.14, while having a larger field of view and a considerably higher number structures included. Conclusion: Our work presents a publicly available full-torso segmentation model for MRI and CT images that classifies almost all subject voxels to date.