Screening for cognitive impairment in older adults’ speech

Artificial intelligence models may be capable of identifying signs of cognitive impairment in the speech patterns of older adults.

In a study published in JAMA Neurology, researchers developed a machine-learning model trained to detect cognitive impairment in the audio recordings of short segments of naturalistic physician-patient interactions during primary care visits. They included 30-second speech segments from nearly 1,000 older participants aged 55 years and older without diagnosed cognitive issues at baseline.

The machine-learning models — particularly those using Whisper-derived acoustic features — were found to be accurate in using acoustic features to predict cognitive impairment. Pitch, timing and variability were the most effective predictors of cognitive impairment. The researchers were able to reproduce the result in an external cohort.

The findings suggested that passive analysis of speech signals could be feasible in helping diagnose cognitive impairment.

Read more: JAMA Neurology

The article presented here is intended to inform you about the broader media perspective on dentistry, regardless of its alignment with the ADA's stance. It is important to note that publication of an article does not imply the ADA's endorsement, agreement, or promotion of its content.