Abstract: An estimated 46.7 million people aged 65 and older have Alzheimer’s or a related form of dementia. Because there is no cure, early detection and prediction of Alzheimer’s is viewed as the best strategy for initiating early treatment and managing the disease. The goal of this study was to demonstrate the potential of using both unstructured telephone conversations and analytics to detect voice features predictive of Alzheimer’s.
The model assessed differences in speech features and accurately identified more than 96% of the
audio files of callers who had diagnostic histories of Alzheimer’s. Voice analysis of telephone conversations is an emerging tool to help predict and detect patients with Alzheimer’s disease and related forms of dementia.
