The University of California, San Francisco (UCSF) was awarded a $1,453,500 Project Grant by the National Institutes of Health (NIH) under the Trans-NIH Research Support program (CFDA 93.310). The funding will support the development of a novel, physiologically-focused approach to train multi-modal artificial intelligence (AI) algorithms for medical applications.
Specifically, the project aims to create a new deep neural network architecture that can accept and learn from multiple inter-related medical data modalities, similar to how a physician triangulates information from various sources to arrive at diagnoses and treatment plans. Additionally, the project will develop an automated data pre-processing and harmonization pipeline to address the heterogeneity of real-world medical data. These innovations will be applied and validated to identify heart failure-related phenotypes in a large cohort of heart failure patients. The expected outcome is a general-purpose AI platform that can enable the development of high-performing, multi-modal medical AI algorithms for a broader range of clinical applications.
Generated 7/16/24, 4:41 AM