Researchers from the University of Toronto have created an artificial intelligence algorithm capable of identifying potential indicators of brain health through everyday speech. In the study, participants were tasked with describing intricate images using their own words. Following this, they underwent standard cognitive assessments while the AI examined their speech patterns, including the frequency and duration of pauses, the use of filler words, the rate of speech, and other characteristics. The analysis revealed a correlation between these speech indicators and the results of cognitive tests, even after accounting for factors such as age, gender, and educational background. It is crucial to note that while this research suggests a connection, it does not imply that dementia can be diagnosed solely based on speech patterns, nor does every pause indicate a dementia risk. Challenges in word recall and interruptions in speech can vary among individuals and may not be exclusive to older adults. Nevertheless, such technological advancements could serve as early warnings that cognitive function is deteriorating more quickly than what is typically expected. Traditional cognitive assessments are often difficult to conduct frequently due to their time-consuming nature, and individuals might enhance their performance due to prior exposure to the tasks. In contrast, speech can be easily recorded in a regular, unobtrusive manner, within a more natural context.
Informational material. 18+.
