Validation of screening method based on EEG analysis for the risk assessment of psychiatric and behavioral disorders: a pilot study

Abstract

To assess the validity of the screening method based on EEG analysis using predictive analytics algorithms with the calculation of linear discriminant functions (LDFs), in comparison with a classification system based on psychometric self-report scales. A comparative cross-sectional study with partial blinding involving healthy volunteers was conducted at two investigational sites. The calculated scores of LDFs used to assess risks of impulsivity, depression and anxiety acted as quantitative characteristics of subjects' mental state. Testing included completing psychometric scales. As a result of the performed validation of the original screening method based on EEG analysis in comparison with the scores of psychometric scales chosen as a reference method, satisfactory results were obtained with the best parameters of sensitivity, specificity, and accuracy for detecting high levels of impulsivity associated with pronounced aggressiveness. Of considerable interest is also the direct correlation found between high levels of LDF impulsivity scores and high levels of self-rated aggression on a psychometric scale (BPAQ-24). The results open up the possibility of using the proposed method to predict a number of emotional and behavioral characteristics of subjects, including a high risk of aggressive behavior as part of professional selection.