Gait foundation model from 3,328 people predicts 1,887 health traits, published in Nature Medicine
segal_eran · x · 2026-11-09
Eran Segal's team published a gait foundation model in Nature Medicine: 3D skeletal motion captured with a single depth camera across 5 motor tasks from 3,328 deeply phenotyped adults trains a self-supervised GaitMAE model. Gait predicts BMI, visceral fat, age, grip strength, sleep quality and 1,887 traits spanning 18 body systems — elevating gait from a symptom of specific pathologies to a systemic biomarker. The pipeline pairs multimodal phenotyping (blood pressure, metabolomics, microbiome, medications) with masked-skeleton reconstruction and hierarchical embeddings for downstream classifiers.
Related event: Gait foundation model predicts 1,887 health traits from walking(2 posts)→
More from Research
- Meta teases entry into AI-for-science race, joining frontier labs push — kimmonismus · 2026-11-09
- Study: 3g Daily Creatine for 4 Weeks Improved Cognition in Women, Most in Older Adults — 2C_ornot2C · 2026-11-09
- Brynjolfsson, Korinek, Agrawal publish Economics of Transformative AI research agenda in CACM — erikbryn · 2026-11-09
- Researchers question AI capability acceleration: shift to RL may be the confounder — DavidSKrueger · 2026-11-09
- Neural network output layers explained: classification vs regression — HarperSCarroll · 2026-11-09
- Ropedia W0 trains robot world models on 10,000 hours of human video, lifting LIBERO-Plus by 6.76 points — jiqizhixin · 2026-11-09