Event Abstract

Using Orthogonal Polynomial Trend Analysis and Wavelet decomposition (WOPTA) to investigate learning in a Mental Rotation task

  • 1 the University of Newcastle, Australia
  • 2 the University of Newcastle, Australia
  • 3 Hunter Medical Research Institute, Australia
  • 4 Monash University, Australia
  • 5 Monash University, School of Psychology & Psychiatry, Australia
  • 6 Monash University, Australia
  • 7 Monash University, School of Psychology & Psychiatry, Australia

BACKGROUND The orthogonal polynomial regression technique (OPTA) has been used previously in paradigms with small trial numbers to investigate learning effects in a habituation ERP paradigm. This analysis approach can dramatically improve SNR. The inclusion of a covariate as part of this regression analysis facilitates investigation of systematic changes in ERP morphology. Traditional ERP analysis methods are blind to the systematic and often dynamic changes that occur during practice or fatigue or in relation to a covariate of interest. Karayanidis et al (2011) used this approach successfully to investigate variability of cue-locked ERP components associated with advance preparation as a function of response time, by applying OPTA in the frequency domain. We are proposing an extension to the OPTA technique, using the continuous wavelet decomposition that will take advantage of the richer time-frequency structure of the data. We are using a mental rotation paradigm designed to encourage transition from mental rotation mastery to direct memory access, using OPTA to investigate the transition between these two strategies.
METHODS We compared a wavelet-based OPTA (WOPTA) and a standard OPTA and against traditional averaging, on both real Mental Rotation data from 11 participants (Shepard & Metzler style stimuli) and simulated data, using time as a covariate. Pre-processing of real EEG data was conducted in Fieldtrip and simulated data was created in BESA simulator using a basic model of EEG data loosely based on neurologically plausible generators as identified by a meta-analysis of mental rotation localisation research.
RESULTS Both covariate approaches were superior to traditional averaging in characterising mental rotation related components of the ERP across time. The added time-frequency resolution of WOPTA over the OPTA technique was able to capture more temporal and frequency specific data whilst largely maintaining cross-frequency coupling and coherency in the ERP data.
CONCLUSIONS WOPTA has the potential to tap into ERP dynamics in a novel way as it not only improves SNR in comparison to averaging, but also reveals the deep covariance structure of the data. It is an improvement to OPTA, which itself has a clear advantage over traditional averaging.

Keywords: WOPTA, wavelet analysis, ERPs, mental rotation, Learning

Conference: ACNS-2013 Australasian Cognitive Neuroscience Society Conference, Clayton, Melbourne, Australia, 28 Nov - 1 Dec, 2013.

Presentation Type: Poster

Topic: Other

Citation: Provost A, Paton B, Karayanidis F, Brown S and Heathcote A (2013). Using Orthogonal Polynomial Trend Analysis and Wavelet decomposition (WOPTA) to investigate learning in a Mental Rotation task. Front. Hum. Neurosci. Conference Abstract: ACNS-2013 Australasian Cognitive Neuroscience Society Conference. doi: 10.3389/conf.fnhum.2013.212.00139

Received: 22 Sep 2013; Published Online: 25 Nov 2013.

* Correspondence: Mr. Alexander Provost, the University of Newcastle, Newcastle, Australia, alexander.provost@newcastle.edu.au

© 2007 - 2014 Frontiers Media S.A. All Rights Reserved