The relationship of hallucinations and psychotic markers

Objective: Hallucinations are included in the symptom cluster of many psychopathologies and are commonly encountered in psychiatry practice as psychotic-like experiences. While some of them are part of the psychotic disorder process within the continuum hypothesis, some of them are included in the non-psychotic disorders diagnostic cluster. Hallucinations present an important field of study with their presence in these two diagnostic clusters, and examination of their formation mechanisms gains importance in terms of diagnosis. There is an abundance of studies on schizophrenia and psychotic disorder in the literature, and the behavior of clinical and psychometric measures in this group is well known. Another clinically important point is the scarcity and insufficiency of objective measurements in diagnostic processes in psychiatry routine. In this field, computerized analysis methods can guide us to gain an objective perspective in the diagnostic process. Based on the characteristics of psychotic disorders that have been defined over the years, this study comparatively examines the role of these markers in non-psychotic disorders. In this way, it was aimed to develop a holistic approach to clinical findings, thought language domain, neurocognition and social cognition dimensions, while at the same time, an objective consistency to this approach was targeted via machine learning algorithms classification analyzes performed on the voice recordings of the participants. Method: Among the participants aged 18-45 included in the study, 28 participants formed the Psychotic Hallucination Group, while there were 29 participants in the Non-Psychotic Hallucination Group. Scale for Assesment of Positive Symptoms (SAPS), Brief Negative Symptom Scale (BNSS), Hamilton Depression Rating Scale (HRSS) were administered to the participants by the researcher. In addition, the Dissociative Experiences Scale (DES), the Borderline Personality Inventory (BPI), the Childhood Trauma Questionnaire (CTQ), and the Cognitive Biases Questionnaire for Psychosis self-evaluation scales were given to the participants. Neurocognitive and Social Cognitive effect was assessed with appropriate psychometric tests. The Bead Task and Babbles Test were applied to assess cognitive biases. Finally, Thought and Language Scale, Non-Semantic Graph Analysis and Acoustic Sound Analysis were applied over the audio recordings taken from the patients in accordance with the DISCOURSE protocol. Classification analyzes were performed with Random Forest Classification Algorithm using graph data and acoustic data. Results: In the analyses, the severity of positive psychotic symptoms was higher in the Psychotic hallucination group. While there was no difference between groups in Amotivation factor, the Expression factor was found to be high in the Psychotic Hallucination Group. Dissociative experiences, borderline personality traits and childhood traumas were found to be higher in the non-psychotic hallucination group in the related self-assessment scales. Neurocognitive performance was similar in both groups, except for the number symbols area. In Social Cognitive tests, while the Psychotic Hallucination group tended to give more hypomentalized responses in theory of mind, hypermentalization was evident in the non-Psychotic Hallucination group. On objective measures, jumping to conclusions bias was higher in the Psychotic Hallucination Group, while it was higher in the Non-Psychotic Hallucination Group on self-assessment scales. In the regression analysis, the predictiveness of clinical data in the diagnosis group was calculated as 84.9%. In classification analysis, the success of Random Forest algorithm was found to be the highest 78,6% for Graph analysis and 75% for Acoustic analysis. Conclusion: In this study conducted with participants with hallucinations, it was seen that hallucinations occur by different mechanisms in psychotic and non-psychotic disorders. In theory of mind, the presence of hypomentalization clinically approximates the diagnosis of Psychotic Disorder, while the presence of hypermentalization was found to be significant for non-psychotic disorders. The difference between objective and subjective cognitive bias measures indicates the importance of metacognition in cognitive functions. The clinical measurements and classification analyzes were found to be consistent with each other and have a high and significant diagnostic predictor. Further studies are needed to develop an objective diagnostic tool.