Alzheimer’s is a neurodegenerative disease that progressively affects memory, language, attention and other essential cognitive functions. It often manifests itself subtly in the early stages. The difficulties may seem minor: longer than usual pauses, reduced vocabulary, simplified grammatical structures or loss of fluency when a person tries to tell a story or describe an experience.
These changes, which may go unnoticed in everyday conversation, are extremely valuable when we know how to analyse them accurately. Spontaneous language is a window into the brain. It allows us to identify the early signs of cognitive decline before they become apparent in other conventional tests.
Speech as a digital biomarker
At Ace Alzheimer Centre Barcelona, we are convinced that language is not only a tool for communication, but also a digital biomarker of the first order. That is why we have developed our own model for the computational processing of spontaneous speech.
This model is capable of capturing these subtle changes. It also analyses them using artificial intelligence techniques. The analysis allows us to translate them into measurable patterns that help differentiate between people at different stages of the disease.
It also allows us to distinguish between different neuropsychological conditions, with the ultimate goal of promoting early detection and improving our understanding of the clinical progression of Alzheimer’s.
Scientific evidence
International research and our own experience confirm this potential. Several recent studies led by Ace have shown that automated speech analysis can accurately classify different stages of the Alzheimer’s continuum, ranging from subjective memory complaints to probable dementia. This classification is possible thanks to the use of acoustic, lexical and syntactic features.
It has also been proven that a brief spontaneous speech protocol can distinguish between unaffected individuals, patients with mild cognitive impairment, and patients with Alzheimer’s dementia. The results show very high accuracy levels. They also allow for the prediction of performance in key cognitive domains such as memory, attention, and executive functions.
A relevant aspect is that acoustic parameters extracted from speech can predict amyloid status in cerebrospinal fluid in people with mild cognitive impairment, a key biomarker in Alzheimer’s research, with greater accuracy than conventional neuropsychological tests.
From scientific research to society
All of this confirms that spontaneous speech not only reflects cognitive status, but also underlying neuropathological processes.
These findings reinforce the idea that spontaneous language, analysed using artificial intelligence, is a powerful and accessible digital marker that allows us to make a qualitative leap in research and early detection of Alzheimer’s disease.
Our commitment as an institution does not end in the scientific field. It also extends to the community. That is why we have promoted projects such as digital screening in community pharmacies. This initiative takes advantage of the proximity and trust offered by neighbourhood pharmacies. It brings technology closer to the public and makes it easier for people over the age of 50 to voluntarily take a short spontaneous speech test using digital tablets.
In these tests, they are asked to describe an image, list words, or answer open-ended questions. The samples collected are analysed using artificial intelligence algorithms that detect subtle speech patterns and indicate a possible cognitive risk. In these cases, they are immediately referred to the Ace Memory Unit for a more comprehensive assessment.
This project is of great value both scientifically and to the community, as it:
- Reduces barriers to early detection.
- Normalises conversation about cognitive health and democratises the use of technology that was previously reserved for specialised clinical settings.
- Consolidates Ace’s role as an institution that brings together research, innovation and social proximity.
- It demonstrates that it is possible to transfer scientific advances to everyday life and make science truly useful to the community.
On Ace’s 30th anniversary, we want to emphasise that what we say and how we say it is much more than a form of communication: it is a reflection of our brain. Analysed with technological precision, spontaneous language helps us detect subtle changes, better understand the disease and move towards more careful, accessible and inclusive early detection of Alzheimer’s.
Dr. Sergi Valero Ventura
Head of Medical Psychology at Ace Alzheimer Centre Barcelona
