Health is one of the main focuses of Apple and its products, and the Apple Watch is one of the devices that the Cupertino company relies most on to monitor users’ vital signs. Ongoing research has now resulted in 95% accuracy when it comes to heart murmurs.
Apple Watch dedicated to users' hearts
Apple has been at the forefront of cardiovascular technology since it launched the Apple Watch about a decade ago. Many of the company's innovations, such as atrial fibrillation detection and irregular heartbeat detection features, have proven to save lives.
A new machine learning research initiative from Apple now shows that it is also possible to detect heart murmurs with great precision.
Heart murmurs are classified based on when they occur in a heartbeat. It can be Systolic, Diastolic or Continuous.
According to Apple's research paper, its MTL model, multi-task learning model (2dCNN-MTL) achieves greater than 95% accuracy in detecting murmurs. This result surpasses existing models, maintaining a MAE (Mean Absolute Error) of 1.636 bpm in estimating heart rate, satisfying the requirements established by the Association for the Advancement of Medical Instrumentation (AAMI).
According to the Cleveland Clinicheart murmurs are common and do not necessarily indicate a health problem, especially in children. However, murmurs should be evaluated, as they may indicate a heart problem or other health problem.
Some heart murmurs, however, may indicate heart-related problems or other critical health issues. These include anemia, carcinoid syndrome, endocarditis, heart valve disease, hyperthyroidism, and hypertrophic cardiomyopathy.
Currently, the doctor listens to the heart with a stethoscope at different points on the chest and back to detect murmurs and prescribes a chest x-ray, an echocardiogram and an electrocardiogram.
Heart murmur detection on Apple Watch, along with the existing ECG app and atrial fibrillation detection, will certainly be another life-saving innovation.
Source: pplware.sapo.pt
