Verily, NVIDIA and NIH launch first multimodal AI model combining EHR and genomics
Verily Health, in partnership with NVIDIA and the NIH All of Us Research Program, announced the development of a new foundation model that merges electronic health record data with genomic information. The effort aims to create a more comprehensive view of patient health for research and clinical use.
The multimodal model represents the first attempt to integrate these two data streams at scale, building on prior AI work that relied solely on clinical records. By combining genetic insights with routine health data, the system seeks to improve disease risk prediction and support more personalized treatment strategies.
Jonathan Amar, senior manager of Verily Data Science, emphasized that electronic health records capture only one chapter of a patient’s story. He noted that incorporating imaging, genomics and other modalities is essential for AI to fully understand disease risk and advance precision medicine.
This writeup was produced by pharmadog from original reporting by BioPharma Dive.
Original headline: “The next frontier: Advancing foundation models with multimodal data”
read at BioPharma Dive ↗
comments(0)
5-min edit window · permanent after that