Applied mathematician and TED Fellow Max Little has been working on ways to diagnose Parkinson's disease by analyzing voice recordings, but now he's turning to the public for help. The Parkinson's Voice Initiative, launched today at the TEDGlobal conference in Edinburgh, is seeking volunteers to submit their voices via a three-minute phone call that will help to refine the algorithms used in research led by Little.

The incurable disease is difficult to diagnose at present due to the lack of biomarkers like blood tests, but almost all patients exhibit vocal tremors that can be picked up by algorithms. Right now the technique has an 86 percent accuracy rate when used in blind testing, rising to 99 percent in mid to late stages of the disease. The new initiative is hoping for 10,000 new samples from people with or without Parkinson's in order to improve those figures. A quick and non-invasive procedure such as voice analysis could be invaluable in providing wide, inexpensive screening at an early stage, and could stop patients having to visit clinics for checkups.

If you'd like to contribute to the Parkinson's Voice Initiative by making a short phone call, there are numbers set up for the US, UK, Canada, Mexico, Spain, Argentina, and Brazil over at the group's website.