The goal of this observational study is to evaluate accuracy of portable electronic stethoscope and machine learning-based diagnostic algorithms for detecting the disease in people with valvular heart disease and healthy controls. The main question it aims to answer is:

Is portable electronic stethoscope and machine learning-based diagnostic algorithms allow accurate detection of valvular heart disease?

Researchers will compare diagnostic algorithm's predictions with the clinicians' predictions to see if the diagnostic results are accurate.

Participants will

• take echocardiogram
• take electrocardiogram using BPM Core
• get the heart auscultation data measured via electronic stethoscope

The study compares the diagnostic accuracy of machine learning-based algorithms for diagnosis, which utilise auscultation data obtained through electronic stethoscopes, with the diagnoses made by clinicians using the same data. Two portable electronic stethoscopes used will be evaluated in this study, including BPM Core (Withings, France) and BeamO (Withings, France). The study utilises data collected from 100 patients at Queen Mary Hospital who have been diagnosed with valvular heart diseases (including aortic stenosis, mitral and/or tricuspid regurgitation, and mitral stenosis) and 25 healthy individuals without heart conditions.

Inclusion Criteria:

• Voluntarily agrees to participate by proving written informed consent
• Have echocardiography done within 5 years

Exclusion Criteria:

• Mechanical heart valve
• Adult congenital heart disease

25 with aortic stenosis, 25 with mitral stenosis, 25 with mitral regurgitation, 25 with tricuspid regurgitation, and 25 without valvular heart disease