The Medical Imaging and Data Resource Center (MIDRC): Building Trustworthy AI through Curated, Harmonized Imaging Data
Description
This session will review the current status and expansion of MIDRC as a unique, large-scale source of curated and harmonized imaging data spanning multiple modalities and body regions. MIDRC’s resources are specifically designed to meet the AI community’s critical needs for high-quality, trustworthy, and diverse datasets. This session will also review the MIDRC BDF Imaging Hub (BIH), a federated platform enabling researchers to search, discover, and analyze biomedical imaging data across multiple repositories at scale. The symposium will also highlight new tools for evaluating data quality and standards (e.g., LOINC), assessing AI algorithm performance, detecting and mitigating bias, and supporting translational research using MIDRC’s sequestered datasets. ). This session will also highlight MIDRC's collaborations with federal agencies, such as the FDA for MDDT submission, and will underscore the central role MIDRC plays in advancing AI for medical physics.