HEALTH

How Dreams and Tech Help Spot Mental Health Issues

ItalyWed Oct 22 2025

People with psychosis often struggle to organize their thoughts and words, making communication challenging. Traditionally, doctors have relied on interviews and questionnaires to assess this. However, a novel approach is emerging: using computers to analyze language.

Study Findings

Researchers analyzed dream reports from 193 Italian speakers, some with psychosis and others healthy. They discovered that individuals with psychosis exhibited more disjointed language patterns, with words appearing less connected, akin to a jumbled puzzle. This trend persisted even after accounting for education levels.

Methodology

The study employed graph theory, a mathematical tool that visualizes connections between elements. In this case, it mapped how words link in speech. People with psychosis showed smaller and weaker connections in their word maps, suggesting less smooth thought flow.

Computer Classification

A computer program classified the dream reports, accurately identifying psychosis 68% of the time. While not perfect, this indicates potential for computers to assist in diagnosing mental health conditions.

Significance

This is the first study of its kind with Italian speakers, contributing to growing evidence that language analysis can aid in understanding psychosis. It offers a fresh perspective on how people think and communicate when unwell.

questions

    How does the Naïve Bayes classifier's accuracy of 68% compare to other machine learning models used in similar studies?
    How does the cross-linguistic validity of graph-based approaches impact the generalizability of these findings to other languages and cultures?
    How do the findings of this study compare with previous research on language disorganization in psychosis using traditional clinical interviews?

actions