Innovative application surpasses doctors in detecting Alzheimer’s disease
Scientists are working on an app that utilises speech analysis to diagnose various conditions
A collaborative effort between researchers at Japan’s University of Tsukuba and engineers from IBM Research’s R&D division has yielded a prototype application capable of early-stage diagnosis of Alzheimer’s disease (AD) and mild cognitive impairment (MCI).
This application utilises machine learning technology to analyse short-term pitch fluctuations, volume changes, speech speed, intonation, and pause frequency. By detecting specific speech patterns, the application achieved an accuracy of 91% in identifying patients with AD and 88% in identifying those with MCI. Early detection of these conditions can significantly slow down their progression, according to the scientists involved.
In a related development, the US National Institutes of Health (NIH) recently committed around $14 million to the development of software that uses artificial intelligence to diagnose various diseases based on the voices of patients. This AI-powered system aims to identify conditions such as Alzheimer’s disease, respiratory diseases, Parkinson’s disease, depression, bipolar disorder, and speech delays in children.
