As a recently retired doc I can comment on the AI tool with some confidence. No doc (and that includes Cardiologists) nowadays would dream of relying on auscultation for much information about heart valves! We listen to hearts sometimes but that is a tiny part of any serious assessment.
If there is any doubt about the heart valves then some combination of echocardiogram, ECG and blood tests would be a minimum.
The reason for doing those tests is seldom that a murmur has been heard. You can of course have severe heart and valve problems without abnormal heart sounds.
In the pre-antibiotic era gross valvular heart disease causing murmurs was common. Now it is pretty uncommon if not rare in developed societies.
The tool may well be better than most docs at finding abnormalities. Most modern ECG machines produce a printout showing abnormalities too. Nearly all these can be ignored on sight as they are of “no clinical importance” - in other words they don’t matter at all. They just worry patients and sometimes cause unnecessary further investigations.
The question is does the tool produce clinically useful results - does it improve patient care?
Biting the hand that feeds IT
