Benefits of functional PCA in the analysis of single-trial auditory evoked potentials

Jan Koláček*, Ondřej Pokora, Daniela Kuruczová, Tzai-Wen Chiu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Evoked potentials reflect neural processing and are widely used to studying sensory perception. Here we applied a functional approach to studying single-trial auditory evoked potentials in the rat model of tinnitus, in which overdoses of salicylate are known to alter sound perception characteristically. Single-trial evoked potential integrals were generated with sound stimuli (tones and clicks) presented systematically over an intensity range and further assessed using the functional principal component analysis. Comparisons between the single-trial responses for each sound type and each treatment were done by inspecting the scores corresponding to the first two principal components. An analogous analysis was performed on the first derivative of the response functions. We conclude that the functional principal component analysis is capable of differentiating between the controls and salicylate treatments for each type of sound. It also well separates the response function for tones and clicks. The results of linear discriminant analysis show, that scores of the first two principal components are effective cluster predictors. However, the distinction is less pronounced in case the first derivative of the response.

Original languageEnglish
Pages (from-to)617-629
Number of pages13
JournalComputational Statistics
Issue number2
StatePublished - 1 Jun 2019


  • Functional data
  • Principal component analysis
  • Single-trial auditory response


Dive into the research topics of 'Benefits of functional PCA in the analysis of single-trial auditory evoked potentials'. Together they form a unique fingerprint.

Cite this