Computer scientists have developed an AI-based speech screening tool that helps assess potential Parkinson’s disease symptoms. Featured in npj Parkinson’s Disease, this web-based test requires users to recite two pangrams, allowing the AI to analyze voice patterns linked to the disease with nearly 86% accuracy. While it’s not a replacement for clinical diagnosis, the tool serves as an accessible method for individuals, particularly in remote areas, to identify the need for further evaluation. Researchers amassed data from over 1,300 participants to train the AI on diverse speech characteristics. Parkinson’s affects nearly 89% of individuals through voice alterations, making it an effective screening factor. The initiative seeks to improve global access to neurological care through common speech interfaces like Amazon Alexa. The study was supported by multiple reputable organizations, aiming to enhance early detection methods for this rapidly growing neurological condition. For more information, an interactive demo of the tests is available online.
