In the past decade, the performance of brain-computer interfaces based on steadystate visual evoked potentials (SSVEPs) has been significantly improved due to advances in signal analysis algorithms. For example, efficient target-identification methods based on template matching, in which individual templates are obtained by averaging the training data across trials, have been proposed to improve the performance of SSVEP detection. In template-based methods, spatial filtering plays an important role in improving the performance by enhancing the signal-to-noise ratio of SSVEPs. In conventional studies, several spatial-filtering approaches have been introduced for electroencephalogram analysis. However, the optimal spatial-filtering approach for individual template-based SSVEP detection still remains unknown. This chapter reviews the spatial-filtering approaches for improving the template-based SSVEP detection and evaluates their performance through a direct comparison using a benchmark dataset of SSVEPs.

Chapter Contents:

• Abstract
• 11.1 Introduction
• 11.2 Individual template-based SSVEP detection
• 11.2.1 Basic framework
• 11.2.2 Ensemble strategy
• 11.2.3 Filter bank analysis
• 11.3 Spatial-filtering techniques
• 11.3.1 Average combination
• 11.3.2 Minimum energy combination
• 11.3.3 Canonical correlation analysis
• 11.3.4 Independent component analysis
• 11.3.5 Task-related component analysis
• 11.4 Material and methods
• 11.4.1 Dataset
• 11.4.2 Performance evaluation
• 11.5 Results and discussions
• 11.5.1 Signal features of SSVEPs after spatial filtering
• 11.5.2 A comparison of frameworks for SSVEP detection
• 11.5.3 A comparison of electrodes settings
• 11.5.4 Toward further improvement
• 11.5.5 Challenges and future direction
• 11.6 Conclusions
• References

Inspec keywords: spatial filters; medical signal detection; visual evoked potentials; brain-computer interfaces; electroencephalography; medical signal processing; signal classification

Other keywords: template matching; individual templates; template-based SSVEP detection; reviews; spatial filtering techniques; signal-to-noise ratio; training data across trials; brain-computer interfaces; efficient target-identification methods; signal analysis algorithms; template-based methods; benchmark dataset; steadystate visual evoked potentials

Subjects: Digital signal processing; Electrical activity in neurophysiological processes; Bioelectric signals; Filtering methods in signal processing; Biology and medical computing; Electrodiagnostics and other electrical measurement techniques