Diagnosis of Schizophrenia and Psychosis in Māori People Using Speech Assay: A Natural Language Processing Approach

Abstract

Indigenous peoples and ethnic minorities worldwide have a higher incidence of psychosis, primarily schizophrenia. In New Zealand, Māori individuals experience greater rates of anxiety and depression compared to non-Māori, with a significantly higher incidence of mental illness among Māori. We propose natural language processing (NLP) trained on speech samples from Māori patients as a potential solution to the problem of culturally biased psychometric screening tools for psychosis and schizophrenia. This research examines NLP's ability to diagnose psychosis in Māori patients by analysing speech and language abnormalities as indicators of severe mental illnesses such as schizophrenia. Our research emphasizes the need for inclusive language models and investigates cross-cultural applicability. We employed a three-part method: conducting clinical interviews, pre-processing data with the Natural Language Toolkit (NLTK) and applying language classifiers. The study's results demonstrate the promise of NLP, but limited patient data necessitates further research, including standardizing datasets and integrating NLP with indigenous languages. This research represents a step towards improving diagnostic accuracy and support for Māori people suffering from psychosis, aligning with healthcare's goal of fair and culturally responsive mental health screening.

Received: 15 March 2024 | Revised: 15 May 2024 | Accepted: 29 May 2024

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

The data that support this work are available upon reasonable request to the corresponding author.