UK Hospitals To Test AI “Death Calculator” For Predicting Patient Life Expectancy

In a bold advancement for healthcare, UK hospitals are set to trial a cutting-edge artificial intelligence system, termed the “death calculator,” to predict life expectancy in patients undergoing standard ECG exams. Known as AI-ECG Risk Estimation, or AIRE, this technology identifies hidden heart-related risks that may escape even the most experienced cardiologists, offering predictions from just minutes of heart monitoring data.

Developed to analyze electrocardiogram (ECG) readings reflecting the heart’s electrical activity AIRE has shown remarkable accuracy. Recent studies suggest the tool can assess a patient’s 10-year mortality risk with up to 78% reliability. Starting mid-next year, two NHS trusts in London will pilot the program, with the aim of extending AIRE’s availability across NHS hospitals within five years.

Dr. Arunashis Sau, a cardiology registrar at Imperial College Healthcare NHS Trust, clarified that AIRE’s purpose is not to replace doctors but to enhance their diagnostic capabilities. “The goal here is to use the ECG as a tool to identify higher-risk individuals who may benefit from further, more detailed tests,” said Dr. Sau. “While ECGs are inexpensive and widely available, AIRE could guide more specific testing, leading to better patient management and potentially reducing adverse events.”

Rather than merely analyzing ECGs for current issues, AIRE dives deeper, identifying subtle electrical and genetic indicators of potential heart conditions. By detecting signs of irregular heart rhythms and potential heart failure, often before symptoms arise, AIRE enables clinicians to take preemptive actions, tailoring treatment to individual patient needs.

“AIRE’s goal was to achieve something beyond human capability, not just expedite what doctors already do but uncover heart structure, function, and genetic factors that a doctor cannot detect in heart tracings,” explained Dr. Sau. Hundreds of patients will participate in the initial trial at Chelsea and Westminster Hospital NHS Foundation Trust alongside Imperial College, generating data to help refine and expand the system for nationwide NHS use.

Research published in Lancet Digital Health today reinforces AIRE’s potential. Drawing from a dataset of over 1.16 million ECGs from around 189,000 patients, the AI model successfully predicted future heart failure in nearly 80% of cases and detected atherosclerotic cardiovascular disease with 70% accuracy.

Dr. Fu Siong Ng, a consultant cardiologist at Imperial College Healthcare NHS Trust, envisions a future where every hospital ECG can inform clinicians about not only immediate diagnoses but also potential future risks, allowing for early intervention. With rising cases of heart attack and stroke among younger populations due to factors like obesity, hypertension, and diabetes, the potential for AI-driven early detection becomes increasingly critical.