A predictive tool for personalised psychosis treatment Sep 02, 2024

Some of the project's UPC researchers, from left to right: Cecilio Angulo, Àngela Nebot, Caroline König and Pedro Jesús Copado. Researcher Alfredo Vellido is also involved.

Approximately 1.2% of the global population is estimated to have psychosis. One of the most common psychotic disorders is schizophrenia, affecting 24 million people worldwide. However, despite effective therapeutic options, only one-third of affected individuals receive specialised treatment.

Within this context, the European project Towards a personalised medicine approach to psychological treatment for psychosis (PERMEPSY) aims to transform psychosis therapy through innovative technologies and personalised therapeutic interventions. For this purpose, the Soft Computing (SOCO) research group and the Knowledge Engineering Research Group (GREC) from the Intelligent Data Science and Artificial Intelligence (IDEAI) research centre at the Universitat Politècnica de Catalunya - BarcelonaTech (UPC) are developing a personalised medicine platform.

Using predictive models created with machine learning techniques, the new tool will predict the most suitable personalised treatment for each patient. The platform integrates retrospective data from 700 psychosis patientstreated with metacognitive training (MCT). Sociodemographic, clinical, cognitive and metacognitive data, and biomarkers associated with treatment response in schizophrenia are collected.

As explained by SOCO researcher and UPC project coordinator Caroline König, "by analysing markers from retrospective data and patient characteristics, the tool will predict the probability that an individual will complete the metacognitive training therapy or not, and also its effectiveness and the possibilities of improvement. This will allow treatment to be personalised for each patient."

The prototype platform will undergo a year-long pilot clinical trial with over 250 patients starting in September. The trial will be conducted by the project's clinical partners, who have collaborated over the years to develop MCT strategies for psychosis patients at different stages of the disorder. The ultimate goal is to provide the global medical community with a free tool to improve psychosis treatment.

The project involves UPC researchers responsible for data analysis and predictive model development, and a software development team.

The consortium also brings together research centres from Spain, Germany, France, Poland and Chile, and is led by the Sant Joan de Déu Research Foundation. The team includes experts in clinical psychology, psychiatry, molecular biology, statistics, computer science and artificial intelligence.