JAMIA open · 2026
Data-derived Identity Verification as a principle for the dissemination, harmonization, and artificial intelligence reuse of sensitive biomedical data.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Cordier Benjamin A, Benton Erik S, Dysinger Eamon J, McWeeney Shannon K, AI-READI Consortium
Аннотация
OBJECTIVES: To facilitate responsible biomedical data sharing, large-scale data harmonization, and cutting-edge artificial intelligence (AI) applications by augmenting controlled-access data dissemination models (DDMs) with Data-derived Identity Verification (DIVe).
MATERIALS AND METHODS: We developed a watermarking-based DIVe system that embeds a cryptographically verifiable and traceable identity within each data file. This system was deployed for the Artificial Intelligence Ready and Exploratory Atlas for Diabetes Insights Consortium and currently watermarks a 3.8TB multimodal dataset for each approved data request.
RESULTS: Over 1.8 petabytes, spanning over 900 data copies and 160 million files, have been uniquely watermarked and disseminated. Private provenance tracing, a novel data sharing license, and educational materials extend trust and accountability beyond controlled environments.
DISCUSSION: Typical controlled-access DDMs assume data remain at rest. Data-derived Identity Verification provides an organizing principle for augmenting DDMs such that data can move while maintaining protection from data leaks.
CONCLUSION: Watermarking-based DIVe supports secure, traceable dissemination of sensitive data, supporting the development of AI technologies in biomedicine.
MATERIALS AND METHODS: We developed a watermarking-based DIVe system that embeds a cryptographically verifiable and traceable identity within each data file. This system was deployed for the Artificial Intelligence Ready and Exploratory Atlas for Diabetes Insights Consortium and currently watermarks a 3.8TB multimodal dataset for each approved data request.
RESULTS: Over 1.8 petabytes, spanning over 900 data copies and 160 million files, have been uniquely watermarked and disseminated. Private provenance tracing, a novel data sharing license, and educational materials extend trust and accountability beyond controlled environments.
DISCUSSION: Typical controlled-access DDMs assume data remain at rest. Data-derived Identity Verification provides an organizing principle for augmenting DDMs such that data can move while maintaining protection from data leaks.
CONCLUSION: Watermarking-based DIVe supports secure, traceable dissemination of sensitive data, supporting the development of AI technologies in biomedicine.