A comparison of classification techniques for the P300 Speller

By Dean J. Krusienski, Eric W. Sellers, François Cabestaing, Sabri Bayoudh, Dennis J. McFarland, Theresa M. Vaughan, Jonathan R. Wolpaw. Journal of Neural Engineering, 2006

843 citations recorded by the source corpus.

This study assesses the relative performance characteristics of five established classification techniques on data collected using the P300 Speller paradigm, originally described by Farwell and Donchin (1988 Electroenceph. Clin. Neurophysiol. 70 510). Four linear methods: Pearson's correlation method (PCM), Fisher's linear discriminant (FLD), stepwise linear discriminant analysis (SWLDA) and a linear support vector machine (LSVM); and one nonlinear method: Gaussian kernel support vector machine (GSVM), are compared for classifying offline data from eight users. The relative performance of the classifiers is evaluated, along with the practical concerns regarding the implementation of the respective methods. The results indicate that while all methods attained acceptable performance levels, SWLDA and FLD provide the best overall performance and implementation characteristics for practical classification of P300 Speller data.

Subfields: Cognitive Neuroscience, Signal Processing

View Paper: https://www.semanticscholar.org/paper/d7c1c9242de85a375032090283144b50116ff6df

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Signal Processing

Citations in the corpus

(C)overt attention and visual speller design in an ERP-based brain-computer interface (2010, methodology): “To choose a symbol in the Matrix speller, the row (out of 6 rows) and the column (out of 5 columns) with maximum classifier outputs were selected, and the target symbol was given by their inter-section [9].”
Flashing characters with famous faces improves ERP-based brain–computer interface performance (2011, methodology): “Classiﬁcation of target versus non-target stimuli was performed using stepwise linear discriminant analysis (Krusienski et al 2006).”
Control of a humanoid robot by a noninvasive brain–computer interface in humans (2008, methodology): “…robot by a noninvasive brain–computer interface in humans In our interface, the subject's command to the robot is determined based on a visually evoked EEG response known as P3 (or P300) [11] [17] [18] which is produced when an object that the user is attending to suddenly changes (e.g., flashes).”
EEG-Based Brain-Computer Interfaces (2017, methodology): “Finally, when alternative translation algorithms are submitted to ofﬂine testing using the same data set, their differences in performance are often minimal [44].”
How many people are able to control a P300-based brain-computer interface (BCI)? (2009, neutral): “Such systems are mostly used as spelling devices because a high number of different characters to choose from can enhance the communication speed of the BCI [7].”
A Noninvasive Brain-Actuated Wheelchair Based on a P300 Neurophysiological Protocol and Automated Navigation (2009, methodology): “The location of the electrodes was selected according to previous P300 studies [26], at FP1, FP2, F3, F4, C3, C4, P3, P4, T7, T8, CP3, CP4, Fz, Pz, Cz, and Oz, according to the international 10/20 system.”
A novel P300-based brain-computer interface stimulus presentation paradigm: moving beyond rows and columns (2010, background): “For example, researchers have explored various electrode montages (Krusienski et al., 2006), stimulus properties such as inter-stimulus interval (ISI) and matrix size (Sellers et al., 2006), and various signal processing methods (Kaper et al., 2004; Krusienski et al., 2006; Lenhardt et al., 2008;…”
P300 brain computer interface: current challenges and emerging trends (2012, methodology): “Krusienski et al. (2006) showed that SWDA and Fisher’s linear discriminant (FLD) provided the best overall performance and implementation characteristics for practical classiﬁcation, as compared to Pearson’s correlation method (PCM), a linear support vector machine (LSVM), and a Gaussian kernel…”
A comprehensive review of EEG-based brain–computer interface paradigms (2019, positive): “Piccione et al [95] extract P300 by using the fuzzy method to combine decomposed components of ICA over EEG. Krusienski et al [96] compared various classification techniques including Pearson ’ s correlation method (PCM); Fisher linear discriminant (FLD); stepwise linear discriminant analysis…”
An auditory oddball (P300) spelling system for brain-computer interfaces. (2009, methodology): “runs two and three were collected in the copy spellingmode of the spelling system (Krusienski et al., 2006; Kübler et al., 2001).”
A brain-computer interface for long-term independent home use (2010, methodology): “SWLDA has been shown to be more effective and computationally simpler than other methods (21).”
A P300-based brain-computer interface for people with amyotrophic lateral sclerosis. (2008, methodology): “The discriminant functions for Phase I offline studies expanded the feature space to include up to 60 spatiotemporal features including signals from nine additional electrodes: F3, F4, C3, C4, CP3, CP4, P3, P4, and Oz. Details of this method are described by Krusienski et al. (2006).”
A review of classification algorithms for EEG-based brain–computer interfaces: a 10 year update (2018, methodology): “However, results presented in [197] suggested that continuous adaption was beneficial for the asynchronous P300-BCI speller, and [227] suggested the same for passive BCI based on the P300.”