The researcher, who had previously been selected as a finalist, states that this recognition represents “a very important boost to consolidate the research line and expand its clinical and social impact”. She also highlights the collective value of the project and the importance of interdisciplinary collaborations.

She also emphasized the role of patients and participants in the study: “This recognition is also for them, who have trusted the project and made it possible to advance in this line of research.”

Speech analysis as a digital biomarker

The project is the researcher’s doctoral thesis, supervised by Dr. Amanda Rodríguez-Urrutia, Dr. Maria Dolores Braquehais, and Dr. Josep Antoni Ramos-Quiroga. The proposal focuses on automatic speech analysis as a digital biomarker for the early detection and monitoring of depression. Using artificial intelligence, natural language processing, and acoustic analysis technologies, the study explores how language patterns, both phonological, semantic, and syntactic, can reflect a person’s emotional state.

This approach could enable the development of fast, objective, and non-invasive tools that complement traditional clinical assessment, particularly useful in settings such as primary care, where consultation time is limited and psychiatric symptoms may be overlooked.

The project is being developed in two phases, which are currently well advanced. On one hand, the team has published a systematic review and meta-analysis in JMIR Mental Health, including 105 studies on the diagnostic accuracy of this type of tools. On the other hand, the first results of a longitudinal clinical study have also been published in the Journal of Affective Disorders, conducted at the Vall d’Hebron University Hospital in collaboration with AcceXible, a start-up that has so far focused on speech analysis models for detecting Alzheimer’s disease and cognitive impairment. The results show good performance of the model in detecting depression, with an AUC of 0.86.

The next step will be to validate these results over time, analyzing whether speech markers can detect changes in symptomatology in 6- and 12-month follow-ups, an aspect that is still little explored in the scientific literature.

Interdisciplinary research with clinical impact

This project is part of a broader research line focused on identifying objective markers and integrating digital technologies into psychiatric practice. It is developed within the Brain-Immune-Gut Unit (BIGU) at Vall d’Hebron, coordinated by Dr. Amanda Rodríguez.

The researcher highlights that what motivates her most is the potential for direct transfer into clinical practice: “This type of research can translate into real improvements for patients, especially in terms of accessibility, personalization, and follow-up of affective disorders”.