ETRI "Early Identification of High-Risk Groups for Dementia Prediction Using 'Artificial Intelligence Technology'"

Artificial intelligence technology is expected to enable early screening of high-risk groups for dementia.

The Electronics and Telecommunications Research Institute (ETRI) announced on the 1st that it is conducting research on artificial intelligence technology that analyzes speech utterances of the elderly to evaluate and predict degenerative brain function decline such as mild cognitive impairment and dementia.

ETRI researchers are explaining the artificial intelligence dementia prediction technology of the elderly voice utterance analysis system. Photo by Korea Electronics and Telecommunications Research Institute

Speech utterance is the result of a complex process in which cognitive functions such as memory, intention, and attention, language generation functions such as phonology, syntax, and semantics, and speech motor functions such as breathing, articulation, and phonation sequentially operate.

In other words, speech analysis makes it possible to early judge and predict the decline of cognitive, language, and motor abilities observed in patients with mild cognitive impairment and dementia.

Based on this, ETRI’s Complex Intelligence Research Lab is expanding its research area to healthcare fields such as digital therapeutics by combining accumulated artificial intelligence technology in speech processing with speech, text, and video multimodal technologies.

In particular, the research team conducted Alzheimer’s dementia prediction research by combining a large language model (LLM) for the first time in the world in addition to existing speech and text analysis technologies, achieving a performance (87.3%) that surpasses the record (85.4%) announced in the ‘ADReSSo Challenge Dataset’ hosted by the University of Edinburgh in the UK and Carnegie Mellon University in the US.

Based on these research results, the team also completed the development of a tablet-based app that predicts high-risk groups for mild cognitive impairment through speech utterance input.

This app was developed with a focus on user convenience and accuracy improvement for actual elderly users, and will undergo verification processes at senior welfare centers in collaboration with the Korea Electrotechnology Research Institute.

The research was conducted as part of the National Research Council of Science & Technology’s creative convergence research project titled “Development of AI-based Degenerative Brain Function Decline Evaluation Technology through Building Big Data of Daily Speech Utterances of the Elderly.”

This content was produced with the assistance of AI translation services.