A group of Spanish researchers created a platform for artificial intelligence for early identification of the disease Alzheimer’s through voice.
The team is made up of members from the University of Alicante (UA) and the Health and Biomedical Research Institute (ISABIAL), both entities in eastern Spain.
The project focused on the detection of Alzheimer’s, since current treatments have been shown to be more effective when administered in the early stages of the disease.
As explained by the UA in a statement, the Artificial Intelligence Platform for the Early Detection of Alzheimer’s Disease through Voice (IAEAV) was designed to identify patterns of cognitive deterioration through the analysis of acoustic and linguistic signals of the human voice.
To do this, they used advanced natural language processing (NLP) and deep learning technologies that allow users, through a mobile application, to record their voices in various contexts, such as reading texts, spontaneous narrations or responses to standardized questions.
Subsequently, the recordings are processed to extract acoustic characteristics, including pitch, intensity and pauses, and linguistic features such as semantic richness and errors in verbal fluency, which are then evaluated by the learning models to obtain a more precise and personalized detection.
The application was designed for use in both clinical and home settings, which, according to the researchers, “reduces access barriers and facilitates obtaining data in populations with limited resources.”
This technology “has not only sought to improve clinical detection, but has also contributed to scientific development by generating large volumes of voice data, which can facilitate deeper investigations into the relationship between linguistic alterations and neurodegenerative changes, promoting advances in the treatment and management of the disease,” indicated one of the researchers.