![[feed]](/style/images/feed-icon-14x14.png){kind=icon}
Tools
Tools
Preview |
PDF (Original Article)
- Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
2MB |
![[thumbnail of Supplementary Information]](https://edoc.mdc-berlin.de/style/images/fileicons/other.png) |
Other (Supplementary Information)
9MB |
| Item Type: | Article |
|---|---|
| Title: | Pooled two-cohort MRI body composition phenotyping with open-source deep learning |
| Creators: |
Mertens, Christian J.  ORCID: https://orcid.org/0000-0001-7303-2650, Häntze, Hartmut ORCID: https://orcid.org/0000-0002-6204-571X, Ziegelmayer, Sebastian, Kather, Jakob Nikolas ORCID: https://orcid.org/0000-0002-3730-5348, Truhn, Daniel ORCID: https://orcid.org/0000-0002-9605-0728, Kim, Su Hwan ORCID: https://orcid.org/0000-0002-5383-2041, Busch, Felix ORCID: https://orcid.org/0000-0001-9770-8555, Weller, Dominik, Wiestler, Benedikt ORCID: https://orcid.org/0000-0002-2963-7772, Graf, Markus, Bamberg, Fabian, Schlett, Christopher L., Weiss, Jakob B., Ringhof, Steffen ORCID: https://orcid.org/0000-0003-1823-1037, Can, Elif ORCID: https://orcid.org/0000-0001-5319-7570, Schulz-Menger, Jeanette ORCID: https://orcid.org/0000-0003-3100-1092, Niendorf, Thoralf ORCID: https://orcid.org/0000-0001-7584-6527, Lammert, Jacqueline, Molwitz, Isabel, Kader, Avan, Hering, Alessa ORCID: https://orcid.org/0000-0002-7602-803X, Meddeb, Aymen, Nawabi, Jawed, Schulze, Matthias B. ORCID: https://orcid.org/0000-0002-0830-5277, Keil, Thomas ORCID: https://orcid.org/0000-0002-9108-3360, Willich, Stefan N., Krist, Lilian, Hadamitzky, Martin, Hannemann, Anke ORCID: https://orcid.org/0000-0003-4420-5449, Bassermann, Florian ORCID: https://orcid.org/0000-0003-4435-2609, Rueckert, Daniel ORCID: https://orcid.org/0000-0002-5683-5889, Pischon, Tobias ORCID: https://orcid.org/0000-0003-1568-767X, Hapfelmeier, Alexander, Makowski, Marcus R. ORCID: https://orcid.org/0000-0001-8778-647X, Bressem, Keno K. and Adams, Lisa C. ORCID: https://orcid.org/0000-0001-5836-4542
|
| Abstract: | BACKGROUND: Body mass index fails to capture variation in fat and muscle distribution that determines metabolic health and disease risk. MRI enables radiation-free quantification of regional body composition, yet scalable open-source tools applied in pooled cohorts with differing acquisition protocols have been lacking. METHODS: MRSegmentator, an open-source nnU-Net-based pipeline, was applied to quantify visceral adipose tissue (VAT), abdominal subcutaneous adipose tissue (ASAT), gluteofemoral adipose tissue (GFAT), trunk musculature, and the liver mask used for liver fat-fraction estimation in 45,851 adults from the German National Cohort (n = 26,877, 3 T multi-centre Siemens) and UK Biobank (n = 18,974, 1.5 T Siemens). Population-scale compartment volumes were segmented from stitched in-phase gradient-echo (GRE) images in both cohorts; liver fat fraction was calculated from fat-only and water-only images. The annotated development data comprised NAKO T2-HASTE and UKB Dixon reconstructions. A single pooled model was applied without site-specific adaptation. A separate two-reader agreement study used 50 scans from these annotated development-sequence domains. Associations between BMI-adjusted body composition and cardiometabolic conditions were estimated using generalized linear mixed-effects models. Incremental discrimination beyond age, BMI, and waist-to-hip ratio was assessed. RESULTS: Five-fold participant-stratified internal cross-validation against curated human-in-the-loop development references comprising UKB Dixon and NAKO T2-HASTE yielded a mean Dice of 0.91. In a separate 50-scan reader study on these annotated development-sequence images, overall reader-reader Dice was 0.937 and overall algorithm-reader Dice was 0.908. The trained pipeline was then used to segment compartment volumes from stitched in-phase GRE inputs in both cohorts, while liver fat fraction was calculated from fat-only and water-only images; direct sequence-matched validation on NAKO GRE was not performed. VAT showed the strongest positive associations with cardiometabolic conditions, while GFAT showed inverse associations, most prominently for type 2 diabetes (OR 0.69, 95% CI 0.66 to 0.72). Disease-specific body-composition phenotypes were identified, with type 2 diabetes characterized by elevated VAT, reduced GFAT, and increased liver fat. MRI-derived compartments modestly improved discrimination for type 2 diabetes and hyperlipidemia beyond anthropometric measures. CONCLUSIONS: A single open-source deep-learning pipeline enabled pooled body-composition phenotyping in two cohorts and captured distributional variation in fat and muscle beyond BMI. High agreement in internal cross-validation (mean Dice 0.91) and the separate two-reader study support the annotated development-sequence analysis, while the population-scale application identified distinct disease-associated phenotypes and modest incremental discrimination beyond conventional anthropometry. |
| Source: | Communications Medicine |
| ISSN: | 2730-664X |
| Publisher: | Springer Nature |
| Volume: | 6 |
| Number: | 1 |
| Page Range: | 467 |
| Date: | 29 August 2026 |
| Official Publication: | https://doi.org/10.1038/s43856-026-01888-w |
| PubMed: | View item in PubMed |
| Related to: |
Repository Staff Only: item control page
