AI Model to Predict Mortality Risk
Researchers have developed an AI tool to help doctors identify high-risk heart patients. Trials in England will begin soon. The AI model, called AI-ECG Risk Estimation (AIRE), predicts early death likelihood after a heart scan.
The research team, led by Imperial College London, trained AIRE using millions of ECG results. An ECG records heart’s electrical activity and helps diagnose heart attacks and other heart problems. The AI aims to detect subtle patterns indicating higher risk of complications or death.
Tests showed AIRE accurately predicted mortality risk within a decade of an ECG with 78% success. Dr. Fu Siong Ng, a cardiac electrophysiology expert at Imperial College, emphasized its potential benefits for the NHS and beyond.
Broader Applications and Future Implementation
Besides predicting death, AIRE forecasts heart attacks, heart failure, and arrhythmias. Plans are in place for NHS-wide use within five years. Trials involving real patients will start by mid-2025 at several London sites. The model’s effectiveness will be tested on patients from outpatient clinics and hospital wards.
Previous AI-based ECG models diagnosed heart diseases but did not assess personal risk levels. Bryan Williams, chief scientific and medical officer at the British Heart Foundation, highlighted AIRE’s potential to transform ECG usage. He believes it could offer better insights into future health risks, including mortality.
The study, published in The Lancet Digital Health, acknowledged possible inaccuracies in AI predictions. These could result from unexpected patient outcomes or unaccounted treatments. Despite this, researchers noted AIRE’s strength in identifying subtle heart changes that may escape doctors’ attention.
Dr. Arunashis Sau, the study’s lead researcher, explained that cardiologists use experience and guidelines to interpret ECGs. In contrast, AIRE spots more subtle irregularities, potentially identifying issues long before they worsen.
More research is needed to clarify AIRE’s role in healthcare settings. The model’s potential impact extends to patients with conditions like diabetes that affect heart health. Dr. Ng believes AIRE could enhance treatment, improving both longevity and quality of life for heart patients.