LeDXA Model Predicts Diabetes and Aging from Scans
iScienceLuvr · x · 2026-08-15
An interesting paper presents LeDXA, a model based on the LeJEPA architecture. Trained on 11,540 DEXA scans, it extracts clinically meaningful whole-body representations from bone density images.
Key findings:
- Disease Prediction: Capable of predicting Type 2 Diabetes and osteoporosis.
- Biological Markers: Can estimate biological age.
- Performance: Outperforms standard scanner-derived DEXA measures and the general-purpose model DINOv3 in cross-cohort prediction, despite using 150,000x fewer training images.
More from Research
- Google-led paper uses sequential auctions to organize decentralized LLM agent workflows — _onionesque · 2026-10-16
- How meloform works: WASM pitch detection plus differentiable circuits learning 1,022 sine coefficients — henloitsjoyce · 2026-10-16
- Since OpenAI's math release, GitHub contributors keep driving down the integer-multiplication kappa — xennygrimmato_ · 2026-10-16
- RFdiffusion3 debuts in Science: generative AI designs proteins and nucleic acids atom by atom — AllThingsApx · 2026-10-16
- New Science report urges NSF to run experiments on science funding reform — alexeyguzey · 2026-10-16
- AI Turns Scrap Impurities into Useful Ingredients for Stronger Recycled Iron — webthing01 · 2026-10-16