Nonlinear analysis of EEG in major depression with fractal dimensions

Abstract

Major depressive disorder (MDD) is a psychiatric mood disorder characterized by cognitive and functional impairments in attention, concentration, learning and memory. In order to investigate and understand its underlying neural activities and pathophysiology, EEG methodologies can be used. In this study, we estimated the nonlinearity features of EEG in MDD patients to assess the dynamical properties underlying the frontal and parietal brain activity. EEG data were obtained from 16 patients and 15 matched healthy controls. A wavelet-chaos methodology was used for data analysis. First, EEGs of subjects were decomposed into 5 EEG sub-bands by discrete wavelet transform. Then, both the Katz's and Higuchi's fractal dimensions (KFD and HFD) were calculated as complexity measures for full-band and sub-bands EEGs. Last, two-way analyses of variances were used to test EEG complexity differences on each fractality measures. As a result, a significantly increased complexity was found in both parietal and frontal regions of MDD patients. This significantly increased complexity was observed not only in full-band activity but also in beta and gamma sub-bands of EEG. The findings of the present study indicate the possibility of using the wavelet-chaos methodology to discriminate the EEGs of MDD patients from healthy controls.

Gamma oscillations as a biomarker for major depression: an emerging topic

“…In addition, subjects performing emotion-related tasks in major depression can show decreased frontal cortex gamma 23,24 . Akar et al 20 found increased resting complexity of gamma signaling in frontal and parietal cortex, in subjects with major depression. Given the task- and anatomy-specific nature of gamma band activity shown in rodent work, it is highly likely that this heterogeneity in the human literature reflects the varying conditions used in different studies.…”

“…The experimental approaches used in these studies vary, with some measuring gamma power during cognitive tasks 18,19 , and others analyzing the oscillatory power spectrum at rest (i.e., baseline) 20–22 . These studies have used various gamma frequency sub-bands, anatomical regions and behavioral states ranging from memory or emotional tasks, to sensory-evoked measures, and not surprisingly have yielded varying results.…”

“…Further testing of the gamma-biomarker hypothesis will also include how gamma rhythms may interact with other candidate biomarkers for depression, including other frequency bands 20,21,59 . Human EEG studies in the past several decades have shown that lower frequency bands, such as theta and alpha, are also promising candidate biomarkers 7–9 .…”

“…One possibility is that gamma measured in different brain regions, perhaps especially when combined with additional biomarkers, may help distinguish these various disorders from one another, or may in general help in detecting their comorbidity. Limited data already suggest, for example, that gamma abnormalities are present in prefrontal cortex in both major depression and schizophrenia 19,20,23,24,72 .…”

Non-lineal EEG Modelling by Using Quadratic Entropy for Arousal Level Classification

“…Indeed, the difficulty to detect changes in physiological EEG features through traditional lineal methods has been reported [17]. Otherwise, EEG recording approach presents some advantages over the rest of biosignals due to the standardization of the electrode placement on the scalp (10-20 system), guaranteeing the reproducibility of the experiments and the comparison among different studies.…”