Characterising psychosis subtypes using eye-tracking and modelling
Research type
Research Study
Full title
Characterising psychosis subtypes using eye-tracking and modelling
IRAS ID
266538
Contact name
Sukhi Shergill
Contact email
Sponsor organisation
King's College London
Duration of Study in the UK
0 years, 5 months, 4 days
Research summary
Summary of Research
Overlapping features among mental disorders make psychiatric diagnosis a challenge. An example of this is psychosis, which can be seen in mental disorders such as schizophrenia and bipolar disorder, and sometimes these disorders can be mistaken for one another. People who have psychosis have impairments in cognitive functions such as decision making, working memory and attention. The impairments in these cognitive domains can be studied using eye-movement behaviour of people with psychosis as the way we explore our environment is underwritten by these cognitive functions. Our objective is to study the differences between psychosis types (e.g., schizophrenia, bipolar disorder) based upon visual exploratory behaviour of patients.In this study, we will use a mathematical model that can predict human visual exploratory behaviour. This model captures distinct cognitive processes that are hypothesised to be impaired in psychosis (e.g., decision making, attention). We will disentangle between psychosis types based on eye movement behaviour and modelling eye movement data. This will help us to quantify the differences among psychosis types. We expect that this study will enhance our knowledge about psychosis types. Characterising psychosis types might lead to more personalised therapy choices for the patients.
We will recruit 35 patients with schizophrenia or any related psychotic spectrum disorder from South London and Maudsley NHS Foundation. We will invite patients to the Institute of Psychiatry, Psychology & Neuroscience at King’s College London. The patients will carry out some questionnaires/tests. We will ask the patients to sit in a chair, rest their head on a forehead- and chin- rest and explore visual scenes with their eyes under different contexts. The eye-movements of the participants will be collected using an eye-tracker as they perform the task. This study will involve one session of around 2-3 hours per patient.
Summary of Results
Objectives:
The main objective of this study was to characterise psychosis subtypes in terms of their eye movements and computational modelling of their visual exploratory behaviour. The secondary objective was to describe the differences between the participants with psychosis and controls in terms of computational modelling and eye movements. Data analysis has been completed to a large extent, and we confirm that the study has achieved its objectives. The future work involves refining the modelling and statistical results.Participants:
The study involving participants with psychosis was given a favourable ethical opinion by the London – Brighton & Sussex Research Ethics Committee. The study involving healthy controls was approved by the King’s College London research ethics committee. We recruited 31 participants with psychosis and 26 healthy controls. Written informed consent was established before participation in the study. The psychosis and control participants were comparable in terms of age.Summary of main findings:
This study explored the cognitive processes guiding the visual exploratory behaviour in psychosis using eye-tracking and computational modelling. The participants performed a series of context-dependent visual exploration tasks. In each task, the scenes involved context-relevant and irrelevant locations, and the participants’ capacity to focus on the context-relevant locations were assessed in terms of behavioural measures. To understand the cognitive processes guiding the participants’ visual exploratory behaviour, we used computational models of the tasks on the fixation locations of each subject. This analysis revealed two psychosis subgroups characterised in terms of model parameters and age, per eye movement task. In each task, one psychosis subgroup explored the visual scenes more efficiently, and their performance was comparable to that of controls. Moreover, a considerable number of participants that performed well on one task also performed well on the other task, and these participants formed the body of the well-performing psychosis subgroups. Surprisingly, there were no significant differences between the relatively well and poorly performing psychosis subgroups in terms of the Positive and Negative Syndrome Scale (PANSS) and Social and Occupational Functioning Assessment Scale (SOFAS) scores. These results suggest that there are distinguishable psychosis subtypes that can be identified in terms of their eye movements, and the ability to use contextual information differs between these subtypes. To address our secondary objective, we computed the differences between participants with psychosis and controls in terms of model parameters. This analysis revealed a decreased capacity to focus on the context-relevant stimuli in the psychosis group relative to the controls.Dissemination of research results:
The research results will be published in peer-reviewed scientific journals. We will send a summary of the research results to the participants who opted to get feedback using their preferred method of contact.REC name
London - Brighton & Sussex Research Ethics Committee
REC reference
19/LO/1437
Date of REC Opinion
31 Oct 2019
REC opinion
Further Information Favourable Opinion