Abstract
In EEG studies, one of the most common ways to detect a weak periodic signal in the steady-state visual evoked potential (SSVEP) is spectral evaluation, a process that detects peaks of power present at notable temporal frequencies. However, the presence of noise decreases the signal-to-noise ratio (SNR), which in turn lowers the probability of successful detection of these spectral peaks. In this paper, using a single EEG channel, we compare the detection performance of four different metrics to analyse the SSVEP: two metrics that use spectral power density, and two other metrics that use phase coherency. We employ these metrics find weak signals with a known temporal frequency hidden in the SSVEP, using both simulation and real data from a stereoscopic apparent depth movement perception task. We demonstrate that out of these metrics, the phase coherency analysis is the most sensitive way to find weak signals in the SSVEP, provided that the phase information of the stimulus eliciting the SSVEP is preserved.
Highlights
While acquiring an EEG signal is easy, the difficulty of recording decent quality signal cannot be overstated: since the data acquisition equipment is working with very small voltages, it is very susceptible to various external and internal noise sources, which are often several times more powerful than the signal intended to be measured
EEG remains popular since it has excellent temporal resolution and it is non-invasive: in the clinic it may be used to detect anomalous oscillations in patients with epilepsy or migraine (Camfield et al, 1978; Adeli et al, 2003), it may be used for characterizing a transient or steady-state response in the brain (Ciganek, 1961; Norcia et al, 2015), or may be used in Brain-Computer Interface (BCI) applications (Bayliss and Ballard, 2000; Nakanishi et al, 2018)
In studies where Fast Fourier Transform (FFT) is used to analyse the spectrum of the state visual evoked potential (SSVEP), an acceptable approach is to take the neighboring 10–20 Fourier components surrounding the temporal frequency of interest, and calculate its average value
Summary
While acquiring an EEG signal is easy, the difficulty of recording decent quality signal cannot be overstated: since the data acquisition equipment is working with very small voltages, it is very susceptible to various external (electrical grid, smartphones, etc.) and internal (eye blinks, vascular pulse, etc.) noise sources, which are often several times more powerful than the signal intended to be measured. Various signal processing techniques can clean up an existing recording only to a certain degree, which may not be sufficient in applications where the signals are exceptionally weak Despite these problems, EEG remains popular since it has excellent temporal resolution and it is non-invasive: in the clinic it may be used to detect anomalous oscillations in patients with epilepsy or migraine (Camfield et al, 1978; Adeli et al, 2003), it may be used for characterizing a transient or steady-state response in the brain (Ciganek, 1961; Norcia et al, 2015), or may be used in Brain-Computer Interface (BCI) applications (Bayliss and Ballard, 2000; Nakanishi et al, 2018). In the 1960s, this was done with presenting flashes of light, and the EEG signal was recorded by taking a photograph of the EEG trace displayed on an oscilloscope (Ciganek, 1961).
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call