AI Tool Helps Pinpoint Problem Heart Cells in Ventricular Tachycardia

Ventricular tachycardia is a potentially life-threatening condition where the rhythm of the heart’s chambers is suddenly disrupted. It is often treated with catheter ablation, in which energy is used to destroy cells causing abnormal impulses. However, accurately pinpointing these cells remains difficult, and more than half of patients relapse within a year because some problem cells remain untreated. Now, artificial intelligence (AI) could help address this challenge.

Researchers at King’s College London (KCL, London, UK), in collaboration with University College London (UCL, London, UK and international partners, have developed an AI tool to identify arrhythmia-causing cells. The tool analyzes complex electrical signals collected during cardiac mapping and detects patterns that may be missed by clinicians. By providing more precise targeting for ablation procedures, it aims to improve patient outcomes and reduce relapse rates.

To test its effectiveness, the team used pig models, as their hearts closely resemble human hearts. Ventricular tachycardia was induced in 13 pigs, and thousands of electrical signals were recorded. Four machine learning algorithms were evaluated, with the random forest model performing best. It identified problem cells with a sensitivity of 81.4% and specificity of 71.4%, showing strong potential for guiding ablation therapy.

The findings, published in European Heart Journal - Digital Health, highlight a proof-of-concept that AI can support cardiologists during ablation procedures, potentially reducing both the risk of relapse and the reliance on broad interventions. The researchers are now refining their approach with more advanced AI methods, including graph neural networks, which have shown encouraging early results. Human studies are already underway to test the model’s clinical utility.

“We envisage this study as a first step towards the development of an automated system that will guide cardiologists in the identification of optimal ablation targets,” said Dr. Michele Orini, Senior Lecturer in Healthcare Engineering at KCL. “Our aim is to improve outcomes for patients, reduce risk and cut the duration and cost of the procedure.”