Machine learning avatars detect signs of dementia with high accuracy


In Japan, a joint group of researchers from Osaka University and Nara Institute of Science and Technology demonstrated that it was possible to detect dementia from conversations in human-agent interaction. As the population ages, an increasing number of people are developing dementia and easy-to-use dementia detection tests are needed.

The researchers proposed machine learning algorithms for detecting signs of dementia in its early stages, by using interactive computer avatars. They created a model based on features of speech, language, and faces from recorded dialogues with elderly participants. A computer came to be able to distinguish individuals with dementia from healthy controls at a rate of 90 percent in 6 questions.

This high accuracy was achieved by combining features of dementia, such as delay in response to questions from avatars depending on the content of questions, intonation, articulation rate of the voice, and the percentage of nouns and verbs in utterance.