Introduction

The invention is an easy to use patient health screening systems, which analyses the human voice, to diagnoses potential or ongoing illnesses and/or diseases.  The human voice is used as a bio-marker, as it changes according to the ailment, in distinguishable and measurable ways.

 

Technical features

The voice of a subject has thousands of parameters (such as fundamental frequency, jitter, shimmer, signal-to-noise ratio, etc.). In healthy conditions, all those parameters are within specific intervals. In the case of illness or disease, some of those parameters (even very few) are outside of their ranges. For different illness or diseases, out range parameter are different, which can be used to establish which illness/disease the subject is suffering from. The aim of this invention is to use voice analysis to investigate the subject’s health/unhealthy conditions, in particular to support diagnosis, for illness or disease in starting, evolving or ongoing conditions. Smartphone apps can be used to record voice samples and via custom analytical algorithms, large numbers of audio features can be analysed to identify and diagnose illnesses. It has been validated in hospitals, for determining Parkinson’s (95% accuracy), Dysphonia (97% accuracy) and Essential Tremor (96% accuracy) and ongoing trials are aim to identifying COVID-19 Infection, Idiopathic Pulmonary Fibrosis, multiple respiratory diseases, sleeping disorders and metabolic surgery effects related to diabetes.

Possible Applications

• Support for medical diagnosis;
• Support for the evaluation of the effectiveness of a drug and of the therapy;
• Tele-assistance, tele-monitoring and tele-medicine.

Advantages

• Applicable to a wide range of pathologies;
• Immediate response;
• Easy to use and low cost;
• Non-invasiveness;
• Possibility of remote investigation.