How can AI help identify early symptoms of Parkinson’s disease (PD)? This is what a recent study published in npj Parkinson’s Disease hopes to address as a team of researchers investigated using an AI-based online tool designed to identify PD through voice patterns. This study has the potential to help researchers, medical professionals, and the public better understand new methods that can be used to identify early symptoms of PD, including appropriate treatment plans.
For the study, the researchers used a combination of speech pattern software (Wav2Vec 2.0 and WavLM) and an AI model (ImageBind) to characterize the speech patterns of 1,306 participants consisting of 392 PD patients, along with 53.2 percent women and 46.8 percent men. The goal of the study was to ascertain the accuracy of using voice patterns for identifying PD. In the end, the tool successfully demonstrated 85.7 percent accuracy for identifying if a participant had PD.
“Research shows that nearly 89 percent of people with Parkinson’s have a deformity in their voice that can be indicative of the disease, making speech a strong starting point for digital screening,” said Tariq Adnan, who is a computer science PhD student at the University of Rochester and lead author of the study. “By combining this method with assessments of other symptoms, we aim to cover the majority of people through our accessible screening process.”
Approximately one million Americans and more than 10 million people worldwide suffer from PD, with men have approximately 1.5 times increased likelihood of being diagnosed with PD compared to women. It is estimated that by 2050 the number of people worldwide that will suffer from PD will be approximately 25,2 million.
How will AI help improve identification of early Parkinson’s disease in the coming years and decades? Only time will tell, and this is why we science!
As always, keep doing science & keep looking up!
Sources: npj Parkinson’s Disease, EurekAlert!, Parkinson’s Foundation