Threshold Determination Criterion in Steady-State Visual Evoked Potential-Based Acuity Assessment: A Comparison of Four Common Methods

Abstract

The steady-state visual evoked potential (SSVEP) visual acuity is usually defined by extrapolating a straight line regressed through significant SSVEP amplitudes plotted versus spatial frequencies to 0 µV or a noise level floor, or the finest spatial frequency evoking a significant SSVEP. This study aimed to compare the performance of the commonly used threshold determination criteria of the extrapolation technique and the finest spatial frequency technique. Visual acuity was measured both by the Freiburg Visual Acuity Test (FrACT) and SSVEP with vertical sinusoidal reversal gratings in ten adults. The extrapolation technique including three methods of linear extrapolation to zero (C1), linear extrapolation to noise level baseline (C2) and linear extrapolation to zero versus log spatial frequency (C3), and the finest spatial frequency technique with significance determination by canonical correlation analysis (CCA) and "OR" operation (C4) were used to determine the SSVEP visual acuity. Bland-Altman method found a pretty good agreement between the SSVEP and FrACT acuity obtained by all the four threshold estimation criteria. One-way repeated-measures ANOVA and Bonferroni post-hoc analysis found that there was no significant difference among visual acuities measured by FrACT and all the four criteria, except for the visual acuity estimated by C1 slightly higher than that of C2, demonstrating that these visual acuity estimating methods had a similar performance in evaluating the visual function. The correlation and agreement between subjective FrACT acuity and objective SSVEP acuity measured by four criteria respectively were all pretty good, demonstrating that all of these four threshold estimation criteria had a good performance in SSVEP visual acuity assessment.

Enhancing Performance of SSVEP-Based Visual Acuity via Spatial Filtering

“…SSVEP visual acuity was defined by the intersection point between the noise level baseline and the regression line extrapolating from the last significant SSVEP peak to the last data point with an SNR higher than the preset SNR level. For the native combination, previous studies have given the recommended value of SNR level, i.e., 1.0 ( Yadav et al, 2009 ; Zheng et al, 2020b ). However, as shown in Figure 6 , CCA and MSI often obtained the higher SNR of SSVEPs than the native combination, especially in high spatial frequencies close to the visual acuity threshold.…”

“…Canonical correlation analysis, a statistical way to measure the underlying correlation between two sets of multidimensional variables, was first used in SSVEP analysis by Lin et al (2007) . Till now, CCA has become the most widely used method in SSVEPs as a result of its effectiveness, robustness, and simple implementation ( Bin et al, 2009 ; Zheng et al, 2020b ; Li et al, 2021 ). Here, CCA finds the weights w y ∈ ℝ N e ×1 and w xf ∈ ℝ 2 N h ×1 to maximize the linear combinations between y = Y wy ∈ ℝ N t ×1 and x = X f w xf ∈ ℝ N t ×1 representing the multichannel SSVEP signals and the SSVEP reference signals.…”

“…Figure 2 shows an example of the tuning curve for the SSVEP visual acuity estimation criterion used in this study. SSVEP amplitude can be plotted versus spatial frequency, and then a regression line can be extrapolated from the last significant SSVEP peak to a noise level baseline ( Zheng et al, 2020b ). The range for the regression line was between the last significant SSVEP peak and the last data point with an SNR higher than the preset SNR level, and the noise level baseline for each visual stimulus was defined as the mean of the noise of the six spatial frequency steps ( Hamilton et al, 2021b ).…”

“…The range for the regression line was between the last significant SSVEP peak and the last data point with an SNR higher than the preset SNR level, and the noise level baseline for each visual stimulus was defined as the mean of the noise of the six spatial frequency steps ( Hamilton et al, 2021b ). Then, the SSVEP visual acuity was defined as the spatial frequency corresponding to the intersection point between the regression line and the noise level baseline ( Zheng et al, 2020b ; Hamilton et al, 2021a ). Besides, the whole diagram of signal processing in this study is shown in Figure 3 .…”

“…On the basis of this idea, several methods of extracting optimal spatial filters to reconstruct source activities from scalp EEG signals have been carried out to enhance the SNR of SSVEPs. For instance, the basic spatial filtering methods [e.g., Laplacian combination ( Friman et al, 2007 ) and common average reference (CAR) ( Zheng et al, 2020d )] and the model-based spatial filtering methods [e.g., minimum energy combination (MEC) ( Friman et al, 2007 ), canonical correlation analysis (CCA) ( Bin et al, 2009 ; Zheng et al, 2020b ; Li et al, 2021 ), and multivariate synchronization index (MSI) ( Zhang et al, 2014a )] have been applied to improve the performance of SSVEPs. However, to date, little is known about whether there is an enhancement of the spatial filtering technique from multielectrode signals on SSVEP visual acuity.…”