A new European Portuguese corpus for the study of Psychosis through speech analysis

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 7298–7304 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 7298 A new European Portuguese corpus for the study of Psychosis through speech analysis Maria Forj´o1, Daniel Neto2, Alberto Abad1, H. Sofia Pinto1, Joaquim Gago3 1INESC-ID/Instituto Superior T´ecnico, University of Lisbon, Portugal 2Servic¸o de Sa´ude da Regi˜ao Aut´onoma da Madeira, Portugal 3Nova Medical School/Centro Hospitalar de Lisboa Ocidental, Lisboa, Portugal maria.beato.forjo@tecnico.ulisboa.pt, danielcarvalhoneto@gmail.com, alberto.abad@inesc-id.pt, sofia@inesc-id.pt, joaquimgago@me.com Abstract Psychosis is a clinical syndrome characterized by the presence of symptoms such as hallucinations, thought disorder and disorganized speech. Several studies have used machine learning, combined with speech and natural language processing methods to aid in the diagnosis process of this disease. This paper describes the creation of the first European Portuguese corpus for the identification of the presence of speech characteristics of psychosis, which contains samples of 92 participants, 56 controls and 36 individuals diagnosed with psychosis and medicated. The corpus was used in a set of experiments that allowed identifying the most promising feature set to perform the classification: the combination of acoustic and speech metric features. Several classifiers were implemented to study which ones entailed the best performance depending on the task and feature set. The most promising results obtained for the entire corpus were achieved when identifying individuals with a Multi-Layer Perceptron classifier and reached an 87.5% accuracy. Focusing on the gender dependent results, the overall best results were 90.9% and 82.9% accuracy, for female and male subjects respectively. Lastly, the experiments performed lead us to conjecture that spontaneous speech presents more identifiable characteristics than read speech to differentiate healthy and patients diagnosed with psychosis. Keywords: Psychosis, Schizophrenia, Speech Analysis, Machine Learning 1. Introduction A study conducted in 2017 found that around 729 mil- lion people in the world live with a mental health disor- der1. These disorders are complex and can take many forms, such as depressive disorder, anxiety disorders, eating disorders and schizophrenia, among others. The use of computational resources to assist during the diagnosis, study and control of these diseases has been developed throughout the years for many of them, like depressive disorder (Cohn et al., 2009). Studies have demonstrated that the use of a computational approach could be a valuable asset to psychiatrists, in particu- lar when focused on characteristics such as facial ex- pressions, speech characteristics and the semantics of speech (Chaika, 1982; Alpert et al., 1997; Yang et al., 2013; Gosztolya et al., 2020). Psychosis is a clinical syndrome characterized by the presence of symptoms such as hallucina- tions, delusions, thought disorders and disorganized speech (Gaebel and Zielasek, 2015). This disorder is characterized by the individual’s distortion and misconception of the surrounding environment, being also commonly associated with feelings of emotional distress and fear. This clinical syndrome can be a consequence of mental illnesses like schizophrenia, physical diseases, substance abuse or stressful life events. 1https://ourworldindata.org/mental-health, accessed on September 20th 2021 Psychotic episodes are time intervals during which …

Content truncated to 500 words