Ibanez-Soria, D. | Search bioRxiv

bioRxiv2018

Echo State Networks Ensemble for SSVEP Dynamical Online Detection

BackgroundRecent years have witnessed an increased interest in the use of steady state visual evoked potentials (SSVEPs) in brain computer interfaces (BCI), SSVEP is considered a stationary brain process that appears when gazing at a stimulation light source.\n\nNew MethodsThe complex nature of brain processes advocates for non-linear EEG analysis techniques. In this work we explore the use of an Echo State Networks (ESN) based architecture for dynamical SSVEP detection.\n\nResultsWhen simulating a 6-degrees of freedom BCI system, an information transfer rate of 49bits/min was achieved. Detection accuracy proved to be similar for observation windows ranging from 0.5 to 4 seconds.\n\nComparison with existing methodsSSVEP detection performance has been compared to standard canonical correlation analysis (CCA). CCA achieved a maximum information transfer rate of 21 bits/minute. In this case detection accuracy increased along with the observation window length\n\nConclusionsAccording to here presented results ESN outperforms standard canonical correlation and has proved to require shorter observation time windows. However ESN and CCA approaches delivered diverse classification accuracies at subject level for various stimulation frequencies, proving to be complementary methods. A possible explanation of these results may be the occurrence of evoked responses of different nature, which are then detected by different approaches. While reservoir computing methods are able to detect complex dynamical patterns and/or complex synchronization among EEG channels, CCA exclusively captures stationary patterns. Therefore, the ESN-based approach may be used to extend the definition of steady-state response, considered so far a stationary process.\n\nHighlightsO_LIWe present a novel SSVEP dynamical detection approach based on ESN.\nC_LIO_LIThis is the first time ESNs are applied to SSVEP based BCI systems.\nC_LIO_LIWe provide experimental validation of proposed methodology.\nC_LIO_LIExperimental results indicate non-stationarity in SSVEP patterns.\nC_LI

bioengineering↗

bioRxiv2017

Occipital tACS bursts during a visual task impact ongoing neural oscillation power, coherence and LZW complexity

Little is known about the precise neural mechanisms by which tACS affects the human cortex. Current hypothesis suggest that transcranial current stimulation (tCS) can directly enhance ongoing brain oscillations and induce long - lasting effects through the activation of synaptic plasticity mechanisms [1]. Entrainment has been demonstrated in in - vitro studies, but its presence in non-invasive human studies is still under debate [2,3]. Here, we aim to investigate the immediate and short-term effects of tACS bursts on the occipital cortex of participants engaged in a change - of - speed detection task, a task that has previously reported to have a clear physiology - behavior relationship, where trials with faster responses also have increased power in {gamma} - oscillations (50 - 80 Hz) [4]. The dominant brain oscillations related to the visual task are modulated using multichannel tACS at 10 and 70 Hz within occipital cortex. We found that tACS stimulation at 10 Hz (tACS 10) enhanced both (8 - 13 Hz) and {gamma} oscillations, in hand with an increase in reaction time (RT) in the change - of - speed detection visual task. On the other hand, tACS at 70Hz desynchronized visual cortices, impairing both phase - locked and endogenous {gamma} - power while increasing RT. While both tACS protocols seem to revert the relationship reported in [4], we argue that tACS produces a shift in attentional resources within visual cortex while leaving unaltered the resources required to conduct the task. This theory is supported by the fact that the correlation between fast RT and high {gamma}- power trials is maintained for tACS sessions too. Finally, we measured cortical excitability by analyzing Event - Related - Potentials (ERP) Lempel - Ziv - Welch Complexity (LZW). In control sessions we observe that lower {gamma} - LZW complexity correlates to faster reaction times. Both metrics are altered by tACS stimulation, as tACS 10 decreased amplitude of the P300 peak, while increasing {gamma}- LZW complexity. To this end, our study highlights the nonlinear cross - frequency interaction between exogenous stimulation and endogenous brain dynamics, and proposes the use of complexity metrics, as LZW, to characterize excitability patterns of cortical areas in a behaviorally relevant timescale. These insights will hopefully contribute to the design of adaptive and personalized tACS protocols where cortical excitability can be characterized through complexity metrics.\n\nAdditional Title Page FootnotesO_LIWe introduce a bursting tACS protocol to study semi-concurrent tACS effects in the visual system and their impact on behavior as measured by reaction time.\nC_LIO_LIBurst 10 Hz tACS (tACS10) applied to the visual cortex entrained {gamma}-oscillations and increased RTs in a change-of-speed detection visual task more than 70 Hz tACS (tACS70) or Control conditions.\nC_LIO_LIBurst tACS10 also decreased amplitude of the P300 peak, while increasing -power and {gamma}-LZW complexity.\nC_LIO_LIPhysiological and behavioral impact of occipital tACS10 and tACS70 was frequency-specific. tACS70 reduced {gamma}-oscillations after 20min of tACS stimulation.\nC_LIO_LICognitive task may determine cortical excitation levels as measured by complexity metrics, as lower {gamma}-LZW complexity correlates to faster reaction times.\nC_LI

neuroscience↗