Method for exploratory cluster analysis and visualisation of single-trial ERP ensembles

• Published in: Journal of neuroscience methods Vol. 250; pp. 22 - 33
• Main Authors: Williams, N.J., Nasuto, S.J., Saddy, J.D.
• Format: Journal Article
• Language: English
• Published: Netherlands 30.07.2015
• Subjects: Algorithms; Brain - physiology; Brain-Computer Interfaces; Cluster Analysis; Computer Simulation; Datasets as Topic; Electroencephalography - methods; Evoked Potentials; Humans; Language; Language Tests; Models, Neurological; Neuropsychological Tests; Principal Component Analysis; Signal Processing, Computer-Assisted; Signal-To-Noise Ratio; Visual Perception - physiology; Empirical Mode Decomposition; k-means clustering; Stability Index; ERP cluster analysis; Genetic Algorithms
• ISSN: 0165-0270; 1872-678X; 1872-678X
• DOI: 10.1016/j.jneumeth.2015.02.007

The validity of ensemble averaging on event-related potential (ERP) data has been questioned, due to its assumption that the ERP is identical across trials. Thus, there is a need for preliminary testing for cluster structure in the data. We propose a complete pipeline for the cluster analysis of ERP data. To increase the signal-to-noise (SNR) ratio of the raw single-trials, we used a denoising method based on Empirical Mode Decomposition (EMD). Next, we used a bootstrap-based method to determine the number of clusters, through a measure called the Stability Index (SI). We then used a clustering algorithm based on a Genetic Algorithm (GA) to define initial cluster centroids for subsequent k-means clustering. Finally, we visualised the clustering results through a scheme based on Principal Component Analysis (PCA). After validating the pipeline on simulated data, we tested it on data from two experiments - a P300 speller paradigm on a single subject and a language processing study on 25 subjects. Results revealed evidence for the existence of 6 clusters in one experimental condition from the language processing study. Further, a two-way chi-square test revealed an influence of subject on cluster membership. Our analysis operates on denoised single-trials, the number of clusters are determined in a principled manner and the results are presented through an intuitive visualisation. Given the cluster structure in some experimental conditions, we suggest application of cluster analysis as a preliminary step before ensemble averaging.