Project Area Leader: Jihad Obeid

Biomedical data is the fuel that enables AI research for the advancement of personalized and transformative healthcare in South Carolina. Despite the exponential growth in biomedical data, the provision of AI-ready data has several challenges.

Real-world data from the electronic health record (EHR) and other data sources, such as omics, environmental, social determinants of health, and remote monitoring devices, is inherently complex and requires significant expertise to analyze and prepare for research. Moreover, the quality of the data is particularly challenging since the source systems are optimized for health care and not research or AI.

When using patient data, researchers must comply with privacy and confidentiality guidelines under ethical and regulatory mandates. Investigators must navigate IRB approvals when appropriate as well as compliance with HIPAA requirements when protected health information is being used.

All ADAPT computing systems housing confidential data must be maintained on the secure segmented network with protected access, authentication methods, and appropriate network defenses.

To build capacity, ADAPT in SC will address biomedical data readiness for AI (e.g., data acquisition, processing, curation, tools, standards, and preparation for AI models). We will leverage the EHR research infrastructure at the CRUs to establish services, expertise, and training to enable AI-ready data delivery.

Goals:

• Establish an infrastructure for the delivery of biomedical data to accelerate AI-enabled research