Data_Sheet_1_Removal of the Sinusoidal Transorbital Alternating Current Stimulation Artifact From Simultaneous EEG Recordings: Effects of Simple Movin.pdf (1.95 MB)

Data_Sheet_1_Removal of the Sinusoidal Transorbital Alternating Current Stimulation Artifact From Simultaneous EEG Recordings: Effects of Simple Moving Average Parameters.pdf

Download (1.95 MB)
dataset
posted on 29.07.2020 by Małgorzata Żebrowska, Piotr Dzwiniel, Wioletta Joanna Waleszczyk

Alternating current stimulation is a promising method for the study and treatment of various visual neurological dysfunctions as well as progressive understanding of the healthy brain. Unfortunately, due to the current stimulation artifact, problems remain in the context of analysis of the electroencephalography (EEG) signal recorded during ongoing stimulation. To address this problem, we propose the use of a simple moving average subtraction as a method for artifact elimination. This method involves the creation of a template of the stimulation artifact from EEG signal recorded during non-invasive electrical stimulation with a sinusoidal alternating current. The present report describes results of the effects of a simple moving average filtration that varies based on averaging parameters; in particular, we varied the number of sinusoidal periods per segment of the recorded signal and the number of segments used to construct an artifact template. Given the ongoing lack of a mathematical model that allows for the prediction of the “hidden” EEG signal with the alternating current stimulation artifact, we propose performing an earlier simulation that is based on the addition of artificial stimulation artifact to the known EEG signal. This solution allows for the optimization of filtering parameters with detailed knowledge about the accuracy of artifact removal. The algorithm, designed in the MATLAB environment, has been tested on data recorded from two volunteers subjected to sinusoidal transorbital alternating current stimulation. Analysis of the percentage difference between the original and filtered signal in time and frequency domain highlights the advantage of 1-period filtration.

History

References

Licence

Exports