BETA: A Large Benchmark Database Toward SSVEP-BCI Application - PubMed Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Jun 23:14:627.
doi: 10.3389/fnins.2020.00627. eCollection 2020.

BETA: A Large Benchmark Database Toward SSVEP-BCI Application

Bingchuan Liu  1 , Xiaoshan Huang  1 , Yijun Wang  2 , Xiaogang Chen  3 , Xiaorong Gao  1
Affiliations

BETA: A Large Benchmark Database Toward SSVEP-BCI Application

Bingchuan Liu et al. Front Neurosci. .

Abstract

The brain-computer interface (BCI) provides an alternative means to communicate and it has sparked growing interest in the past two decades. Specifically, for Steady-State Visual Evoked Potential (SSVEP) based BCI, marked improvement has been made in the frequency recognition method and data sharing. However, the number of pubic databases is still limited in this field. Therefore, we present a BEnchmark database Towards BCI Application (BETA) in the study. The BETA database is composed of 64-channel Electroencephalogram (EEG) data of 70 subjects performing a 40-target cued-spelling task. The design and the acquisition of the BETA are in pursuit of meeting the demand from real-world applications and it can be used as a test-bed for these scenarios. We validate the database by a series of analyses and conduct the classification analysis of eleven frequency recognition methods on BETA. We recommend using the metric of wide-band signal-to-noise ratio (SNR) and BCI quotient to characterize the SSVEP at the single-trial and population levels, respectively. The BETA database can be downloaded from the following link http://bci.med.tsinghua.edu.cn/download.html.

Keywords: brain-computer interface (BCI); classification algorithms; electroencephalogram (EEG); frequency recognition; public database; signal-to-noise ratio (SNR); steady-state visual evoked potential (SSVEP).

PubMed Disclaimer

Figures

Figure 1
Figure 1
The QWERT virtual keyboard for a 40-target BCI speller. (A) The layout of a conventional keyboard with ten numbers, 26 alphabets and four non-alphanumeric keys (dot, comma, backspace <, and space _) aligned in five rows. The upper rectangle is designed to present the input character. (B) The frequency and initial phase of each target are encoded using the joint frequency and phase modulation.
Figure 2
Figure 2
Typical SSVEP features in the temporal, spectral, and spatial domains. (A) Time course of average 10.6-Hz SSVEP of nine parietal and occipital channels (Pz, PO3, PO5, PO4, PO6, POz, O1, Oz, and O2). The dash line represents stimulus onset. (B) The topographic maps of SSVEP amplitudes at frequencies in the range from the fundamental signal (10.6 Hz) to the fourth harmonic (21.2, 31.8, and 42.4 Hz). The leftmost scalp map indicates the spectral amplitude at the fundamental frequency before stimulus. (C) The amplitude spectrum of the SSVEP of the nine channels at 10.6 Hz. Up to five harmonics are visible in the amplitude spectrum. The averaged spectrum across channels is represented in the dark line in (A,C).
Figure 3
Figure 3
The amplitude spectrum as a function of stimulus frequency (frequency range: 8–15.8 Hz; frequency interval: 0.2 Hz). The spectral response of SSVEP decreases rapidly as the number of harmonics increases and up to 5 harmonics are visible from the figure.
Figure 4
Figure 4
Normalized histogram of narrow-band SNR (A) and wide-band SNR (B) for trials in the benchmark database and BETA. The red diagram indicates the BETA, and the blue diagram indicates the benchmark database.
Figure 5
Figure 5
The wide-band SNR corresponding to the 40 stimulus frequencies (from 8 to 15.8 Hz with an interval of 0.2 Hz). A general declining tendency of SNR with the stimulus frequency can be observed. The SNR is higher at 15.8 Hz presumably because the target has a larger shape of the region.
Figure 6
Figure 6
The phase as a function of stimulus frequency (A) and the bar plot of estimated latencies (B). The SSVEP of Oz channel at nine consecutive stimulus frequencies (row 1 of the keyboard) is extracted for the purpose of analysis. The error bar indicates the standard deviation.
Figure 7
Figure 7
The average classification accuracy (A) and ITR (B) for six supervised methods (msTRCA, TRCA, m-Extended CCA, Extended CCA, ITCCA, and L1MCCA). Ten data lengths ranging from 0.2 to 3 s with an interval of 0.2 s were used for evaluation. The gaze shift time used the calculation of ITR was 0.55 s.
Figure 8
Figure 8
The average classification accuracy (A) and ITR (B) of five training-free methods (FBCCA, CVARS, MEC, MSI, and CCA). Ten data lengths ranging from 0.2 to 3 s with an interval of 0.2 s were used for evaluation. The gaze shift time used the calculation of ITR was 0.55 s.
Figure 9
Figure 9
The scatter plot of narrow-band SNR vs. ITR (A) and wide-band SNR vs. ITR value (B). The dash line indicates a linear model regressed on the data (A: adjusted R2 = 0.393, p < 0.001; B: adjusted R2 = 0.549, p < 0.001). The regression indicates that the wide-band SNR correlated better with the ITR than the narrow-band SNR.
Figure 10
Figure 10
The distribution of wide-band SNR and its fitting to a normal distribution. An individual's SNR of the SSVEP is rescaled to the range of normal distribution by Equation (7) to obtain the BCI quotient.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Abu-Alqumsan M., Peer A. (2016). Advancing the detection of steady-state visual evoked potentials in brain–computer interfaces. J. Neural Eng. 13:036005. 10.1088/1741-2560/13/3/036005 - DOI - PubMed
    1. Allison B. Z., McFarland D. J., Schalk G., Zheng S. D., Jackson M. M., Wolpaw J. R. (2008). Towards an independent brain–computer interface using steady state visual evoked potentials. Clin. Neurophysiol. 119, 399–408. 10.1016/j.clinph.2007.09.121 - DOI - PMC - PubMed
    1. Ang K. K., Guan C. (2013). Brain-computer interface in stroke rehabilitation. J. Comput. Sci. Eng. 7, 139–146. 10.5626/JCSE.2013.7.2.139 - DOI
    1. Bakardjian H., Tanaka T., Cichocki A. (2010). Optimization of SSVEP brain responses with application to eight-command brain–computer interface. Neurosci. Lett. 469, 34–38. 10.1016/j.neulet.2009.11.039 - DOI - PubMed
    1. Bin G., Gao X., Wang Y., Li Y., Hong B., Gao S. (2011). A high-speed BCI based on code modulation VEP. J. Neural Eng. 8:025015. 10.1088/1741-2560/8/2/025015 - DOI - PubMed