AI beyond LLMs: a wearable foundation model based on JEPA

Would a wearable model like this gain in predictive power by adding FHIR/EHR inputs? CoMET [1] has great AUC even without daily/lifestyle metrics like steps, awakenings at night, or vo2max estimates (which are present in say apple health). It would be interesting to know how much you would gain from having a holistic model that sees both inputs.

[1] https://arxiv.org/pdf/2508.12104v1

It's a cool idea! This is going to become easier in the next year since TEFCA will let patients request their own medical records through the health information exchanges that have already been set up for treatment.

Amazing work - as a clinician, being able to have wearable data be clinically useful would be game changing for our patients and for models of healthcare delivery.

(OP here) Happy to answer any questions on this work!