Researchers from the University of Edinburgh in the UK have developed a tool that uses artificial intelligence (AI) to merge patient data with blood test results for NT-proBNP heart protein levels to help doctors detect heart failure earlier.

Named CoDE-HF, the tool leverages AI to merge routinely obtained patient data with NT-proBNP test results to predict if they have suffered heart failure.

Researchers from the university and 13 other countries merged findings from 10,369 individuals with suspected acute heart failure to develop the tool, which could improve clinicians’ decisions and enhance patient care.

CoDE-HF was also highly accurate in hard-to-diagnose patient populations, such as the elderly and people with pre-existing medical conditions.

At present, the team is carrying out additional studies to analyse how this decision-support tool could work in the hospital setting and impact patient outcomes.

Testing if NT-proBNP levels are below a specific cut-off value is the existing diagnosis method used, however, it is not used extensively, as levels can change depending on age, weight and other health conditions.

A life-threatening condition, acute heart failure occurs when the heart is suddenly incapable of pumping blood around the body.

Symptoms, including shortness of breath and swelling of the legs, are observed in various other ailments, which makes diagnosis hard.

University of Edinburgh British Heart Foundation Cardiology professor Nicholas Mills said: “The application of AI in decision-support tools as CoDE-HF to deliver more personalised care is particularly important given our ageing patient population who are living longer with more pre-existing medical conditions. 

“We are currently conducting further studies to identify ways to implement CoDE-HF effectively in routine care.”