Classifying Schizotypy Using an Audiovisual Emotion Perception Test and Scalp Electroencephalography

Area of Science:

• Neuroscience
• Psychiatry
• Machine Learning

Background:

• Schizotypy, a personality trait with "psychotic" symptoms, is linked to schizophrenia spectrum disorders.
• Individuals with schizotypy and schizophrenia patients exhibit emotional processing deficits.
• Early identification of schizotypy is crucial for timely intervention.

Purpose of the Study:

• To investigate multimodal emotion perception in schizotypy.
• To determine if electroencephalogram (EEG) signals can differentiate schizotypy (ST) from neurotypical controls (NC) using machine learning.
• To explore the clinical implications of this method for early risk detection.

Main Methods:

• Forty-five participants (30 ST, 15 NC) underwent an audiovisual emotion perception test with EEG recording.
• Machine learning techniques, including mean subsampling and shrinkage linear discriminant analysis, were applied to EEG data.
• Discriminatory features were extracted for classification.

Main Results:

• The machine learning model achieved over 98% accuracy in distinguishing the schizotypy group from the control group.
• A zero rate of false-positive results was observed.
• The classification method demonstrated high efficacy in differentiating the groups.

Conclusions:

• Machine learning analysis of EEG signals can effectively discriminate individuals with schizotypy.
• This technique holds potential for early clinical identification of individuals at subtle risk for schizotypy.
• Early detection may facilitate timely interventions for schizophrenia spectrum psychopathology.