App identifies disease

VoiceDiag can detect Parkinson’s faster

A NEW smartphone app developed by Melbourne researchers claims to detect Parkinson’s disease in just 10 seconds.

The screening app records the user’s voice and can identify the illness quickly, enabling early diagnosis and treatment.

The changes in a person’s voice are the main indicators of Parkinson’s.

Researchers from RMIT University are using artificial intelligence to conduct the screening test.

Lead researcher Professor

Dinesh Kumar, from RMIT’s School of Engineering, said that the first phase of clinical trials of the VoiceDiag app had been completed with more than 70 per cent accuracy.

“It identifies people with Parkinson’s disease very well.

“It is not something we recommend as a diagnostic at the moment. It is a screening tool or an assisting tool.

“One of the one of the main advantages of our app is that it is a non-language-dependent so anyone from around the world can use it.’’

The technology can also help detect severe Covid-19 because lung infections change a patient’s voice.

Parkinson’s Victoria estimates that about 27,000 Victorians are living with the condition – and almost one in five of those people are still of working age.

Diagnosis for Parkinson’s is currently usually done by a neurologist’s evaluation test, which takes up to 90 minutes. The degenerative brain condition can affect young people and symptoms include slow movement, tremor, rigidity and imbalance.

The cause is unknown. “Early detection, diagnosis and treatment could help manage these illnesses, and so making screening faster and more accessible is critical,” Professor Kumar said.

The test could can be performed anywhere in the world, and would be helpful where doctors monitor patients remotely, he said.

The research results are being published in IEEE Journal of Translational Engineering in Health and Medicine.

Three symptoms – rigidity, tremor and slowness – alter the voice of people with Parkinson’s and these are now identified by the app technology.

“We used voice recordings of people with Parkinson’s and a controlled group of so-called healthy people saying three sounds – A, O and M – which is similar to the Hindu meditation chant,” Professor Kumar said. “These sounds result in a more accurate detection of the disease.”

The research team is looking for a commercial partner and a clinical affiliate before further trials next year.

The RMIT researchers partnered with the Technical University of Kosice in Slovakia, the University of Surabaya in Indonesia and Rajshahi University of Engineering and Technology in Bangladesh on this work.