A groundbreaking study from Imperial College London reveals a new artificial intelligence (AI) tool, CardioKG, poised to dramatically speed up the identification of potential heart disease treatments. This development is crucial given that cardiovascular diseases (CVDs) are the leading cause of death in the European Union, claiming approximately 1.7 million lives each year and impacting 62 million people.
How CardioKG Works: Combining Data for Faster Insights
CardioKG combines detailed heart scans from thousands of participants in the UK Biobank with extensive medical databases. This allows researchers to pinpoint genes linked to heart conditions like atrial fibrillation, heart failure, and heart attacks – accelerating the drug discovery process. The key lies in using “knowledge graphs” : integrated databases that connect information on genes, drugs, and diseases.
According to Declan O’Regan, group leader at Imperial College London, “One of the advantages of knowledge graphs is that they integrate information about genes, drugs and diseases.” By merging imaging data with this knowledge, CardioKG can predict which medicines might be most effective for specific heart conditions with greater accuracy.
Unexpected Drug Candidates and Personalized Care
The AI model highlighted several surprising potential treatments. It suggested that methotrexate, a common rheumatoid arthritis drug, could benefit patients with heart failure, and that gliptins (diabetes medications) might help those with atrial fibrillation. Intriguingly, the analysis also indicated a possible protective effect of caffeine in some atrial fibrillation cases, though researchers caution against changing caffeine intake based on these findings.
The ultimate goal is personalized care : tailoring treatments to an individual’s unique cardiac function. Researchers believe this technology is adaptable to other conditions as well, including brain disorders and obesity.
Looking Ahead: Dynamic Patient-Centric Frameworks
The team plans to expand CardioKG into a “dynamic, patient-centered framework” that tracks disease progression in real time. Khaled Rjoob, the study’s first author, explains this will “open new possibilities for personalized treatment and predicting when diseases are likely to develop.” This next phase promises to leverage AI not just to find drugs, but to anticipate and prevent heart disease before it strikes.
In essence, CardioKG represents a major leap forward in precision medicine for cardiovascular health, offering a faster, more targeted approach to drug discovery and treatment.
