Baycrest, Toronto, York Link Speech Patterns to Dementia Risk With 1-Time AI Analysis
Updated
Updated · BuzzFeed · May 26
Baycrest, Toronto, York Link Speech Patterns to Dementia Risk With 1-Time AI Analysis
3 articles · Updated · BuzzFeed · May 26
AI analysis of recorded picture descriptions found that filler words, pauses and word-finding trouble tracked with poorer cognitive-test performance, suggesting speech could flag higher dementia risk.
The study points to language as an early cognitive domain affected by dementia, but outside experts said occasional forgotten words and slower speech can also reflect normal aging rather than disease.
One major limitation is that the data came from a single time point, making it hard to tell whether a person's use of "um" or long pauses is a lifelong habit, cultural pattern or true decline.
Experts said more concerning signs include repeated short-term memory lapses, getting lost in familiar places and frequent misplacing of items, while risk may be lowered through blood-pressure control, exercise, sleep and heart-healthy habits.
We all forget words. How can new AI tell if it's normal aging or early-onset dementia?
Your phone could soon detect dementia. Is this a medical breakthrough or a privacy nightmare?