Edinburgh Napier University

  Traditional diagnostic tools like the stethoscope have long helped clinicians detect signs of illness through direct contact and active listening. In an era of ubiquitous sensing, the diagnostic potential of devices already in common use, smartphones, hearing aids, earbuds, and voice assistants, deserves further attention. This paper proposes a framework for passive, edge-processed audio monitoring that could assist general practitioners and carers in early detection and condition monitoring, without compromising privacy or requiring new hardware. Use cases include sleep apnoea, cot death risk, respiratory irregularities, and changes in vocal or movement patterns linked to neurological conditions. The concept is offered freely as open prior art, to encourage ethical development and public benefit.