← brief for 2026-09-03 · Slowing decline

Development and external validation of a contrastive learning foundation model for ECG-based prediction of cardiovascular diseases and outcomes

Key takeaways

Researchers built an artificial intelligence model that reads heart tracings (ECGs) to predict heart disease, then tested it on patients from a separate hospital system. This is a prediction study on existing records, so it shows the model spots a pattern, not that it changes outcomes. The target here is heart disease, not kidney disease, so it falls outside this digest.

Who did this work

Michael Ko (Scripps Research Institute) · Matteo Gadaleta (Scripps Research Institute) · Eric J. Topol (Scripps Research Institute) · Evan D. Muse (Scripps Research Institute) · Giorgio Quer (Scripps Research Institute)

Source

Paper · OpenAlex, CKD works · 2026-09-01
https://doi.org/10.1016/j.landig.2026.101092