Multi-modal Modeling in Precision Medicine: From Data Imputation to Synthetic Data

Join the NIH for an introduction to multi-modal data modeling, an emerging approach that leverages large, pre-trained AI foundation models to infer missing biomedical data and create realistic synthetic samples. The session will highlight how multi-modal modeling enables researchers to capture complex relationships across diverse data types, such as omics and imaging, supporting efficient, in-silico investigation of disease processes while reducing reliance on expensive and time-intensive data collection.

What You’ll Learn:

• How cross-modal data modeling uses one data type (like imaging) to fill in gaps in another data type (like genomics).
• Ongoing multi-modal modeling efforts in spatial omics, digital pathology, and radiology.
• How multi-modal modeling is anticipated to help us better understand disease biology and improve healthcare practices.

Visit the event page for more information.