IEEE Computer Society Newsfeed

Subscribe RSS

« Back

Voice Algorithms Applied to Parkinson’s Screening

British mathematician Max Little has developed a noninvasive, inexpensive screening method for rapidly identifying Parkinson’s disease using algorithms that analyze voice recordings. The system evaluates factors such as tremor, breathiness, and weakness. In a blind test, his system accurately detected Parkinson’s sufferers 86 percent of the time. He is training the system to detect when voice anomalies are the result of factors such as colds or smoking, rather than Parkinson’s. Little said he wants the technology to be available to doctors within two years. “We’re not intending this to be a replacement for clinical experts. Rather, it can very cheaply help identify people who might be at high risk of having the disease and for those with the disease, it can augment treatment decisions by providing data about how symptoms are changing between check-ups with the neurologist,” he told the BBC. He said his system could also determine the efficacy of Parkinson’s drugs in clinical trials and, ultimately, help reduce health care costs. People in Brazil, Spain, the UK, and US, among other areas, interested in contributing voice samples can call local numbers to contribute a voice sample to the screening database. Little will present his research at the TEDGlobal conference this week in Edinburgh. (BBC)(Parkinson's Voice Initiative)

Trackback URL: