Apple’s Machine Learning Research can now detect Heart Murmurs with 95% accuracy
Apple has been on the forefront of cardio tech since it rolled out the Apple Watch close to a decade ago. Many of the company’s innovations, such as atrial fibrillation detection and irregular heart-beat detection features have proven to be life saving.
According to Apple’s research paper, their MTL model, multi task learning model (2dCNN-MTL) achieves accuracy over 95% in murmur detection, surpassing existing models, while maintaining an MAE of 1.636bpm in heart rate estimation, satisfying the requirements stated by Association for the Advancement of Medical Instrumentation (AAMI).
According to Cleveland Clinic, Heart murmurs are common and don’t necessarily indicate a health problem, especially in children. However, murmurs should be evaluated because they may signal a heart problem or other health issue.
Some heart murmurs, however, can indicate heart-related issues or other critical health issues. These include anemia, carcinoid syndrome, endocarditis, heart valve diseases, hyperthyroidism, and hypertrophic cardiomyopathy.
