JMIR Publications

ABSTRACT

The rapid proliferation of artificial intelligence (AI) in clinical medicine contrasts sharply with the persistent underrepresentation of low- and middle-income countries (LMICs) in AI datasets. Current FDA-approved AI tools predominantly utilize data from high-income nations, with fewer than 4% reporting geographic or racial diversity. This geographic skew leads to significant performance degradation when used in LMIC populations, perpetuating an equity crisis where health burdens are highest. Addressing this disparity, we highlight Medical Imaging Datasets for India (MIDAS) initiative as a viable model to transition LMICs from "data poverty" to "data sovereignty." MIDAS employs a rigorous, four-domain Dataset Quality Matrix to ensure representativeness, documentation, technical fidelity, and governance, creating openly benchmarked, gold-standard datasets tailored to local contexts. Initial releases, including datasets for oral and dural lesions, demonstrate feasibility and practical value for developing robust, generalizable AI models. Further, we propose a multilateral South-South Data Commons structured around three foundational pillars: a harmonized dataset-grading rubric, distributed custodial governance, and outcome-linked incentives. This infrastructure supports local stewardship, encourages global collaboration, and ensures quality benchmarks drive financial incentives for dataset expansion and diversity. This proposed framework not only positions LMICs as autonomous data stewards but also enhances global AI equity. By institutionalizing quality control, interoperability, and outcome accountability, LMICs can transform from passive data consumers into active contributors of essential, trustworthy, and globally relevant biomedical datasets.