Published on: Dec 29, 2025
Researchers at the MRC Laboratory of Medical Sciences and Imperial College London have developed CardioKG, a new AI-powered technology that combines heart imaging data with biological databases to identify genes and potential drugs linked to heart disease, as reported in Nature Cardiovascular Research. By integrating detailed cardiac imaging into a knowledge graph for the first time, the tool significantly improves the accuracy of predicting disease-associated genes and opportunities for drug repurposing.
Led by Dr Khaled Rjoob and Professor Declan O’Regan, the team trained CardioKG using heart-imaging data from more than 9,500 UK Biobank participants both healthy individuals and patients with atrial fibrillation, heart failure, or heart attack generating over 200,000 imaging traits. These were combined with data from 18 biological databases, enabling the model to uncover new gene–disease links and identify existing drugs with potential cardiovascular benefits, including methotrexate for heart failure and gliptins for atrial fibrillation. The study also revealed a possible protective effect of caffeine in certain atrial fibrillation patients.
Beyond cardiology, CardioKG serves as a proof of concept for applying imaging-enhanced knowledge graphs to other organs, such as the brain or adipose tissue, opening new avenues for research in conditions like dementia and obesity. The researchers aim to further develop the platform into a dynamic, patient-centred system to support personalised medicine and earlier disease prediction, while offering pharmaceutical companies a faster, more targeted approach to drug discovery.
Source: https://lms.mrc.ac.uk/new-ai-driven-tool-could-help-find-heart-disease-drugs-faster/
Back to News
Empowering Innovation Through Global Conferences and Events.
© 2025 SciInov. All Rights Reserved.