New AI systems can detect often overlooked heart disease.
Recently, doctors and scientists at the Smidt Heart Institute at the Siddhasenay Medical Center published an article in JAMA Cardiology that they have for the first time developed an artificial intelligence tool that can effectively identify and distinguish two often overlooked but life-threatening heart diseases: hypertrophic cardiomyopathy and cardiac amyloidosis.
Cardiac amyloidosis, often referred to as "cardiac sclerosis syndrome," is a disease caused by the deposition of abnormal proteins in heart tissue. As amyloid accumulates, it replaces healthy heart muscle, making it difficult for the heart to function properly because patients may have no or occasional symptoms, and it is difficult to detect.
Hypertrophic cardiomyopathy is a disease that causes myocardium to thicken and harden. Its symptoms are usually manifested by the inability of the heart to relax and congestion, which in turn leads to damage to the heart valves, effusion in the lungs, and arrhythmias.
"Both heart diseases are a challenge even for cardiologists, and patients often have years or even decades to get a proper diagnosis," said David Ouyang, MD, a cardiologist at the Smit Heart Institute and senior author of the study. ”
Artificial intelligence detects the principle of heart disease
The new algorithm is applied to more than 34,000 ultrasound videos of the heart. Against these clinical images, the algorithm identified specific features associated with heart wall thickness and ventricular size, effectively labeling certain patients as suspected patients with underlying unrecognized heart disease.
"The new algorithm identifies high-risk patients more accurately than the eyes of trained clinical experts." "This is because the algorithm extracts subtle clues from ultrasound video, thus distinguishing it from benign heart conditions that normally look very similar," Ouyang said. ”
Cardiologists have mentioned that without comprehensive testing, it is difficult to distinguish between similar disease conditions and detect abnormal changes in the shape and size of the heart that are sometimes considered part of normal aging — amyloidosis of the heart and hypertrophic cardiomyopathy, although two different diseases, look similar on echocardiograms.
In the early stages of the disease, both heart diseases can even mimic the appearance of a normal heart and change its size and shape as we age. The algorithm can distinguish between potentially life-threatening heart diseases and send early warning signals before the disease progresses to a point where it may affect health.
Ouyang said they hope the technology can be used to identify patients early in their course, because in order for patients to get the most benefit from existing treatments, the sooner the disease is confirmed, the better, which can be very effective in preventing the worst possible outcomes, such as heart failure, hospitalization and sudden death.
Source | Delve into science
Edit | Chen Bowen