Artificial intelligence identifies individuals with prediabetes using single-lead electrocardiograms
This study establishes that an AI model, DiaCardia, can accurately identify individuals with prediabetes from an ECG alone, with performance that is robust across different patient cohorts and independent of major clinical confounders. Our highly generalizable, single-lead DiaCardia model offers a promising solution for scalable prediabetes screening via wearable devices, potentially enabling early, home-based detection and transforming diabetes prevention strategies.
Authors
- Chikara Komiya (ORCID: https://orcid.org/0000-0002-5605-7335)
- A. Takeuchi (ORCID: https://orcid.org/0000-0002-7246-8644)
- Masato Horino
- Kazunari Hara (ORCID: https://orcid.org/0000-0001-7778-7866)
- Ryo Kaneda (ORCID: https://orcid.org/0000-0002-2035-8814)
- Kazutaka Tsujimoto
- Daisuke Koga (ORCID: https://orcid.org/0000-0002-2792-7207)
- Ryoko Ishii
- Hideki Katagiri (ORCID: https://orcid.org/0000-0002-2786-8569)
- Masanori Murakami (ORCID: https://orcid.org/0000-0002-5541-3502)
- Jun Aoki (ORCID: https://orcid.org/0000-0002-3412-0073)
- Hideyuki Shimizu (ORCID: https://orcid.org/0000-0002-1349-9483)
- Satoshi Ohno
- Kenji Ikeda (ORCID: https://orcid.org/0000-0003-1380-4983)
- Taku Yamada (ORCID: https://orcid.org/0000-0003-2043-4626)
- Rei Okazaki
Institutions
- Tokyo Medical and Dental University (JP)
- Tohoku University (JP)
- Systems Biology Institute (JP)
- Research Organization of Information and Systems (JP)
Publication Details
- Journal
- Cardiovascular Diabetology
- Published
- 2025-11-11
- DOI
- https://doi.org/10.1186/s12933-025-02982-4
- Citations
- 5
- Primary Topic
- ECG Monitoring and Analysis
- Type
- article
- Field-Weighted Citation Impact
- 6.94
Funders
- Japan Science and Technology Agency
- Institute of Medical Science, University of Tokyo
- Moonshot Research and Development Program