An AI screening tool has shown promise in detecting early signs of Parkinson’s disease by analysing speech patterns, potentially supporting earlier diagnosis.
Scientists at the University of Rochester developed the algorithm to detect subtle speech changes linked to Parkinson’s – a progressive neurological condition that affects movement and coordination.
The system was trained on voice recordings from more than 1,300 people, including 392 with Parkinson’s. It achieved nearly 86 per cent accuracy across different sexes, ethnicities and age groups.
Participants took part in a web-based assessment, recording two pangrams – short sentences that contain every letter of the alphabet.
Recordings were made at home, in clinical centres, and at Parkinson’s care facilities.
Tariq Adnan is a PhD student in the Rochester lab and one of the study’s first authors.
The researcher said: “Research shows that nearly 89 per cent 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.”
The tool is not intended to replace formal diagnosis but could act as a fast screening option, particularly in settings with limited clinical resources.
Researchers noted that the algorithm still shows higher error rates in certain subgroups and is being refined.
“This approach not only facilitates convenient at-home monitoring but also plays a crucial role in the early detection and managing the progression of PD [Parkinson’s], potentially altering the course of the disease by enabling earlier therapeutic intervention,” the researchers said.
While most people with Parkinson’s experience speech changes, this is not universal.
The Rochester team has been developing digital screening tools over the past decade, using combined indicators such as motor tasks and facial expressions.
Adnan said: “By combining this method with assessments of other symptoms, we aim to cover the majority of people through our accessible screening process.”
