Artificial intelligence can predict the risk of Alzheimer's disease through speech
Scientists from Boston University have developed a new artificial intelligence algorithm that analyzes the speech of people with mild cognitive impairment (MCI).
Research showed that the algorithm can predict the transition from MCI to Alzheimer's disease within six years with an accuracy of 78.5%. The algorithm was trained on transcripts of audio recordings of 166 people aged 63 to 97 years. Since scientists already knew which participants developed Alzheimer's disease, the computer model was able to find patterns in their speech that indicated a likely decline in cognitive functions. Age and gender, which the participants indicated themselves, were also added to the final prediction.
Computer scientist Ioannis Pashalidis from Boston University explained: "You can think of this indicator as the probability that a person will remain stable or progress to dementia. We wanted to predict what will happen over the next six years, and found that we can do this with sufficient accuracy. This demonstrates the power of AI."
Even if there is no complete cure for Alzheimer's today, early detection offers several advantages: it allows for earlier therapy, studying disease progression, and engaging people in clinical trials of new drugs.
Pashalidis added: "If it is possible to predict what will happen, there are more opportunities and time for intervention with medication to maintain stability and prevent progression to more severe forms of dementia."