Log In
Monday 26th August 2019

Voice algorithms to spot Parkinson's

25th June 2012

Mathematician Max Little has created a program which uses algorithms to detect Parkinson's disease from voice recordings.


Mr Little, who is a TED Fellow, is due to present his findings at the TEDGlobal Conference in Edinburgh and is asking for people to volunteer in order to create a database of recordings. 

At present, detecting Parkinson's disease can take a long time as it cannot be identified using a blood test.

The algorithm detection system has proved to be 86% accurate in tests and Mr Little wants to collect 10,000 voices in order to create a "voice bank".

While Mr Little was doing his PhD at Oxford University in 2003 he was intrigued by how voices could be understood using mathematics.

He told the BBC: "I was looking for a practical application and I found it in analysing voice disorders, for example when someone's voice has broken down from over-use or after surgery on vocal cords."

"I didn't occur to me at the time that people with Parkinson's and other movement disorders could also be detected by the system." 

The algorithm system looks for differences in voice patterns in order to detect the disease.

Mr Little added: "This is machine learning. We are collecting a large amount of data when we know if someone has the disease or not and we train the database to learn how to separate out the true symptoms of the disease from other factors." 


Share this page


There are no comments for this article, be the first to comment!

Post your comment

Only registered users can comment. Fill in your e-mail address for quick registration.

Your email address:

Your comment will be checked by a Healthcare Today moderator before it is published on the site.

Mayden - Innovative cloud-based web development for the healthcare sector
© Mayden Foundation 2019