A computer algorithm, created using artificial intelligence (AI), has demonstrated remarkable accuracy in distinguishing genuine heart attacks from false alarms. This groundbreaking development, according to a study published in Nature Medicine, could potentially enable doctors, particularly those in crowded emergency rooms, to focus their attention on the most critical cases.
Developed by researchers at the University of Edinburgh, the AI tool, named CoDE-ACS, was tested on over 10,000 patients across six countries. It was found to be 99.6% accurate in ruling out patients who were not suffering from heart attacks, showing consistent performance across different ages and genders.
At present, heart attacks are generally diagnosed by measuring the levels of a blood protein called troponin, which is released when heart cells die. However, this traditional method doesn’t account for individual patient differences. Notably, women often present lower levels of troponin even during a heart attack, a nuance that the standard measurement tends to overlook.
The CoDE-ACS algorithm innovatively addresses this issue. It combines protein level data with personalised information such as age, sex, electrocardiogram results, and medical history to calculate the probability of a severe cardiac event.
Cardiology professor Nicholas Mills, who led the research, emphasised the complexity of diagnosing heart disease. He noted that acute chest pain, a key indicator of heart disease, can be caused by various conditions, making diagnosis a complex task.
In a statement, Professor Mills said, “Harnessing data and artificial intelligence to support clinical decisions has enormous potential to improve care for patients and efficiency in our busy emergency departments.”
The rapidly evolving field of AI technology is steadily gaining traction in healthcare. Its applications range from streamlining drug development to enhancing direct patient care. Moreover, pharmaceutical companies are investing heavily in these technologies to expedite the discovery of new drugs.
Some experts even foresee potential applications for AI programs such as ChatGPT in medical education, patient management, and other areas.
The ability to swiftly rule out heart attacks could potentially reduce hospital admissions, thereby easing the pressure on overstretched emergency departments. Further research, including clinical trials in Scotland, is underway to explore this potential.
This innovative integration of artificial intelligence in healthcare, particularly in critical areas like cardiology, promises significant improvements in patient outcomes and hospital efficiency. As further advancements are made, the medical community and patients alike stand to benefit greatly from these technological leaps.