1 2 INVESTIGATING THE ROLE OF ALPHA AND BETA RHYTHMS 3 IN FUNCTIONAL MOTOR NETWORKS 4 ALKINOOS ATHANASIOU, a,b * MANOUSOS A. KLADOS, a 5 CHARIS STYLIADIS, a NICOLAS FOROGLOU, b 6 KONSTANTINOS POLYZOIDIS b AND 7 PANAGIOTIS D. BAMIDIS a 8 a Lab of Medical Physics, School of Medicine, Faculty of 9 Health Sciences, Aristotle University of Thessaloniki (AUTH), 10 54124 Thessaloniki, Greece 11 b First Department of Neurosurgery, AHEPA University 12 General Hospital, 54634 Thessaloniki, Greece 13 Abstract—It is recognized that lower electroencephalo- graphic (EEG) frequencies correspond to distributed brain activity over larger spatial regions than higher frequencies and are associated with coordination. In motor processes it has been suggested that this is not always the case. Our objective was to explore this contradiction. In our study, seven healthy subjects performed four motor tasks (execu- tion and imagery of right hand and foot) under EEG record- ing. Two cortical source models were defined, model A with 16 regions of interest (ROIs) and model B with 20 ROIs over the sensorimotor cortex. Functional connectivity was calculated by Directed Transfer Function for alpha and beta rhythm networks. Four graph properties were calculated for each network: characteristic path length (CPL), clustering coefficient (CC), density (D) and small-world-ness (SW). Different network modules and in-degrees of nodes were also calculated and depicted in connectivity maps. Analysis of variance was used to deter- mine statistical significance of observed differences in the network properties between tasks, between rhythms and between ROI models. Consistently on both models, CPL and CC were lower and D was higher in beta rhythm net- works. No statistically significant difference was observed for SW between rhythms or for any property between tasks on any model. Comparing the models we observed lower CPL for both rhythms, lower CC in alpha and higher CC in beta when the number of ROIs increased. Also, denser networks with higher SW were correlated with higher number of ROIs. We propose a non-exclusive model where alpha rhythm uses greater wiring costs to engage in local information progression while beta rhythm coordinates the neurophysiological processes in sensorimotor tasks. This article is part of a Special Issue entitled: Neurofeed- back. Ó 2016 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: brain waves, electroencephalography, functional connectivity, motor imagery, motor network, sensorimotor cortex. 14 15 INTRODUCTION 16 The performance of physical motor tasks (also known as 17 motor execution) involves the activation and 18 communication of many cortical regions. The mental 19 rehearsal or execution of a motor task that does not 20 actually lead to physical execution is often referred to as 21 motor imagery (Decety and Ingvar, 1990). Motor imagery 22 has drawn a lot of attention especially due to the similar 23 (to motor execution) activation patterns that it elicits in 24 the human brain (Pfurtscheller and Neuper, 1997; 25 Avikainen et al., 2002; Ja¨rvela¨inen et al., 2004) and it 26 has been extensively utilized in rehabilitation practices 27 and sports training. More importantly, motor imagery 28 has been employed in cases of severe neurological dis- 29 ability (such as those caused by strokes or spinal cord 30 injury) as a control modality of Brain-Computer Interfaces 31 (BCIs) to promote communication or functional mobility 32 restoration (Wolpaw et al., 2002; Birbaumer, 2006). 33 Recently, interest has been drawn to the functional 34 characteristics of cortical regions, especially the way 35 that each region communicates with each other and the 36 neurophysiological observations that actually better 37 represent their activation and communication patterns. 38 In a number of electroencephalographic (EEG) studies, 39 neural oscillations within the range of alpha (8–12 Hz) 40 and beta (13–30 Hz), as well as gamma activity (30– 41 90 Hz), have been identified as the EEG bandwidth 42 more commonly associated with the sensorimotor 43 processes (Neuper et al., 2006; Sabate et al., 2011, 44 2012; Lopes da Silva, 2013). Alpha rhythm when 45 recorded over the sensorimotor regions is also known 46 as the mu or sensorimotor rhythm and alpha modulation 47 has drawn attention with regards to its physiological role 48 in motor execution and motor imagery (Sabate et al., 49 2011). An attentional role to both local neuronal http://dx.doi.org/10.1016/j.neuroscience.2016.05.044 0306-4522/Ó 2016 IBRO. Published by Elsevier Ltd. All rights reserved. * Correspondence to: A. Athanasiou, Lab of Medical Physics, School of Medicine, Faculty of Health Sciences, Aristotle University of Thessaloniki, PO Box: 376, 54124 Thessaloniki, Greece. Tel: +30- 2310-999310, mobile: +30-6947-811621. E-mail addresses: athalkinoos@auth.gr (A. Athanasiou), klados@cbs.mpg.de (M. A. Klados), cstyliadis@auth.gr (C. Styliadis), nforoglou@auth.gr (N. Foroglou), polyzoik@auth.gr (K. Polyzoidis), bamidis@med.auth.gr (P. D. Bamidis). Abbreviations: BCIs, Brain-Computer Interfaces; CC, clustering coefficient; CCD, Cortical Current Density; CPL, characteristic path length; D, density; DTF, Directed Transfer Function; EEG, electroencephalographic; FME, foot motor execution; FMI, foot motor imagery; HME, hand motor execution; HMI, Hand motor imagery; MNI, Montreal Neurological Institute; MRI, Magnetic Resonance Imaging; MVAR, multivariate autoregressive; PCC, Pearson’s Correlation Coefficient; ROIs, regions of interest; SW, small-world-ness. Neuroscience xxx (2016) xxx–xxx Please cite this article in press as: Athanasiou A et al. Investigating the role of alpha and beta rhythms in functional motor networks. Neuroscience (2016), http://dx.doi.org/10.1016/j.neuroscience.2016.05.044 1 NSC 17130 No. of Pages 17 31 May 2016