Down syndrome's brain dynamics: analysis of fractality in resting state

Abstract

To the best knowledge of the authors there is no study on nonlinear brain dynamics of down syndrome (DS) patients, whereas brain is a highly complex and nonlinear system. In this study, fractal dimension of EEG, as a key characteristic of brain dynamics, showing irregularity and complexity of brain dynamics, was used for evaluation of the dynamical changes in the DS brain. The results showed higher fractality of the DS brain in almost all regions compared to the normal brain, which indicates less centrality and higher irregular or random functioning of the DS brain regions. Also, laterality analysis of the frontal lobe showed that the normal brain had a right frontal laterality of complexity whereas the DS brain had an inverse pattern (left frontal laterality). Furthermore, the high accuracy of 95.8 % obtained by enhanced probabilistic neural network classifier showed the potential of nonlinear dynamic analysis of the brain for diagnosis of DS patients. Moreover, the results showed that the higher EEG fractality in DS is associated with the higher fractality in the low frequencies (delta and theta), in broad regions of the brain, and the high frequencies (beta and gamma), majorly in the frontal regions.

Keywords: Down syndrome; EEG; Fractal dimension.

Fig. 1

Circles with the double edges show the loci with the FDs significantly different in the DS and normal groups (p value < 0.0027)

Fig. 2

EEG power spectrum at representative frontal, central and parietal midline electrodes for the normal group (represented in a, c, and e, respectively, at the left side, depicted by black lines) and the DS group (represented in b, d, and f, respectively, at the right side, depicted by gray lines)

Fig. 3

Mean values of KFD in each channel for the DS group (depicted by circles) and the normal group (depicted by squares)

Fig. 4

Mean values of KFD in each channel in the intervals 1 (triangle) and 2 (inverted triangle) in the DS and normal groups (int: interval)

Fig. 5

Mean values of KFD in each channel for the DS group (depicted by circles) and the normal group (depicted by squares) at EEG subbands a delta, b theta, c alpha, d theta, and e delta