New study shows how AI could transform drug prescriptions for heart diseases

A new study says an artificial intelligence tool could help speed up the search for heart-disease treatments and make prescribing more personalised. Scientists at Imperial College London built a system called CardioKG, a “knowledge graph” that links detailed heart scans with large medical databases to spot connections between genes, diseases and medicines. It was trained using heart-imaging data from thousands of UK Biobank participants, including people with atrial fibrillation, heart failure and past heart attacks, as well as healthy volunteers. Researchers say this approach can improve how accurately they predict which drugs might help specific heart conditions and could support more tailored care based on how a person’s heart is functioning. Among the medicines highlighted, the model suggested methotrexate could help in heart failure, while gliptins might benefit atrial fibrillation. It also hinted at a possible protective effect of caffeine in some cases, but the team stressed people should not change their caffeine intake based on this.