engleski

EEG characteristics in patients with affective disorder

The aim of the study was to identify specific characteristics of EEG recordings that could be used as potential biomarkers for major depressive disorder or bipolar disorder. For that purpose we have included 30 healthy participants whose EEG recordings were compared with 30 patients diagnosed with major depressive disorder and 10 patients diagnosed with bipolar disorder, both according to the ICD-10 criteria and all of which did a native EEG recording on the day of the admission in the hospital. Our findings confirm some of the previous studies, validating the importance of some characteristics of theta and gamma waves in depression detection, but also disputed some earlier studies regarding relative wavelet energy. Furthermore, our research suggests that wave entropy in beta band wave on the Cz position is a good independent predictor of depression as well as wave entropy of delta band wave on position F7, which also had high predictive accuracy. We did not find any reliable predictors for bipolar disorder. Alongside these findings, we employed machine learning methodology in correlating clinical characteristics and EEG findings.

affective disorder ; EEG ; depressive disorder ; bipolar disorder ; machine learning ; wave entropy

nije evidentirano