Researchers Use Voice Data and AI For Early Diagnosis of Parkinson’s

Parkinson’s disease affects more than 10 million people across the globe, and while there is still no cure, the disease can be better controlled if symptoms are detected early on. One of the main aspects of Parkinson’s is that as it progresses, speech changes.

This led to Rytis Maskeliūnas, a Lithuanian researcher from Kaunas University of Technology (KTU), and a team of colleagues from the Lithuanian University of Health Sciences (LSMU), to set out and try to identify these early symptoms using voice data.

Detecting Subtle Changes in Speech Patterns

According to Maskeliūnas, as motor activity decreases, so does the function of the vocal cords, diaphragm, and lungs.

“Changes in speech often occur even earlier than motor function disorders, which is why the altered speech might be the first sign of the disease,” Maskeliūnas says.

Professor Virgilijus Ulozas at the Department of Ear, Nose, and Throat at the LSMU Faculty of Medicine says that patients with early-stage Parkinson’s might speak in a quieter manner, which is oftentimes monotonous, less expressive, slower, and more fragmented. With that said, these changes are hard to detect by ear.

The joint team of researchers developed a new system that works to solve this problem.

“We are not creating a substitute for a routine examination of the patient — our method is designed to facilitate early diagnosis of the disease and to track the effectiveness of treatment,” says Maskeliūnas.

He also says that the link between the disease and speech abnormalities is not new, but as technology advances, it becomes easier to extract more insightful information from speech.

Using AI Algorithms and Voice Data

Utilizing cutting-edge artificial intelligence (AI), the researchers conducted a groundbreaking study to create tailored analyses and diagnoses of spoken signals in the Lithuanian language. In mere seconds they were able to expand existing AI databases with results unique to Lithuania’s linguistic peculiarities.

Kipras Pribuišis is a lecturer at the Department of Ear, Nose, and Throat at the LSMU Faculty of Medicine.

“So far, our approach is able to distinguish Parkinson’s from healthy people using a speech sample,” Pribuišis says. “This algorithm is also more accurate than previously proposed.”

By recording the speech of both healthy and Parkinson’s patients in a soundproof booth, the team utilized an AI algorithm to process signals. This innovative approach required no complex hardware and could eventually be applied onto mobile devices – paving the way for enhanced care solutions in future.

“Our results, which have already been published, have a very high scientific potential. Sure, there is still a long and challenging way to go before it can be applied in everyday clinical practice,” says Maskeliūnas.

Maskeliūnas has identified the next steps in his research: expanding patient numbers to gain further evidence, comparing this algorithm with other early Parkinson’s detection methods and confirming its efficacy across various contexts such as medical offices or within home environments.