Did You Know...

style="float: right; margin-bottom: 10px; font-weight: 600;"Fri 26th Oct, 2012

A breakthrough technique could detect Parkinson's disease simply by using voice recordings from a thirty second phone call.

Parkinson's disease is a degenerative neurological disorder that affects around 6.5 million people throughout the world. The disease causes the degeneration of dopaminergic neurons which results in depleted levels of dopamine - the neurotransmitter that controls movement. Typical symptoms of the disease include slowness, rigidity, postural instability and the classic 'tremor at rest' that we associate with the disease.


Unfortunately there currently exists no cure or any reliable, definitive method of detection for the disease as the degeneration of the brain cells is not visible through any imaging techniques and there are no bio-markers which will show up in blood tests. Detection therefore is only possible through standardised medical tests administered by specialised neurologists. In Europe alone, 13.9 billion Euro is spent each year in the management of Parkinson's Disease. Some part of that expense would include these costly and resource demanding assessments.


Recent research as part of the Parkinson's Voice Initiative, however, shows promising in-roads to developing a cheap, simple and objective tool for the detection, monitoring and potential screening of the disease. Applied mathematician Max Little from MIT and the University of Oxford is leading the team that is testing a method by which they detect Parkinson's disease by using voice recordings from a thirty second phone call self-administered by the patient at home. "Just as the limbs are affected in Parkinson's, so too are the vocal organs" Dr Little explains, "and we see all the same symptoms; vocal tremor, weakness and rigidity and the speech actually becomes quieter and more breathy after a while". The method sees participants make prolonged "aaaah" sounds over the phone and algorithms are then applied to the data using the latest precision voice analysis software.

In the samples collected to date, this method has been able to detect healthy and Parkinson's voice signals to a level of 99% accuracy. These voice signal tests are high speed, ultra low cost and logistically very feasible, all of which makes them highly scalable. "We can make population scale screening feasible for the first time" explains Dr Little, "and we have the opportunity to start to search for the early biomarkers of the disease before it's too late".


Write a comment ...
Post comment
Cancel