A prototype of a digital application driven by AI has been engineered by a team of graduates from Sechenov University’s Digital Department. This tool is designed to assess the risk of cardiovascular and metabolic ailments using vocal biomarkers.
To initiate the analysis, the user simply needs to record a brief audio file using their smartphone, and the outcome is delivered in a matter of seconds. This information was shared with “Gazeta.Ru” by the university’s press service.
The application systematically analyzes characteristics of the voice signal to gauge associated risks. According to the project’s creators, the human voice reflects the condition of several physiological systems.
Speech engages the respiratory system, the laryngeal muscles, and the articulatory apparatus, alongside underlying neural regulation algorithms. When inflammation, metabolic disruptions, or substantial strain on the circulatory system occurs, voice characteristics shift. While these alterations are virtually undetectable by the unaided ear, they can be identified through digital signal processing and machine learning techniques.
The developers state that the platform is capable of examining vocal patterns to spot predisposition toward arterial hypertension, ischemic heart disease, type 2 diabetes mellitus, and other cardiometabolic conditions.
The creators envision that this technology could eventually be adapted for preliminary risk stratification during medical check-ups, epidemic forecasting, remote patient monitoring, and fulfilling numerous other pressing medical requirements.
Scientific evidence already substantiates the correlation between changes in voice and the status of various bodily organs and systems. However, the achieved accuracy metrics still necessitate independent clinical verification. Issues surrounding the security of personal voice data and ethical considerations also remain to be addressed.
About the Author
