Single-trial analysis using wavelet-based time-frequency filtering, ICA-based spatio-temporal filtering and multiple linear regression analysis
Brief radiant laser pulses can be used to activate cutaneous Aδ and C nociceptors selectively and elicit a number of transient brain responses in the ongoing EEG (N1, N2 and P2 waves of laser-evoked brain potentials, LEPs). Despite its physiological and clinical relevance, the early-latency N1 wave of LEPs is often difficult to be measured reliably, because of its small signal-to-noise ratio (SNR), thus producing unavoidable biases in the interpretation of the results. Here, we aimed to develop a robust method to enhance the signal-to-noise ratio of the N1 wave and measure its peak latency and amplitude in both average and single trial waveforms. We obtained four main findings. First, we suggest provide quantitative evidence that the N1 wave can be better detected using a central-frontal montage (Cc-Fz), as compared to the recommended temporal-frontal montage (Tc-Fz). Second, we show that the N1 wave is optimally detected when the neural activities underlying the N2 wave, which interfere with the scalp expression of the N1 wave, are preliminary isolated and removed using independent component analysis (ICA). Third, we show that after these N2-related activities are removed, the signal-to-noise ratio of the N1 wave can be further enhanced using a novel approach based on wavelet filtering. Fourth, we provide quantitative evidence that a multiple linear regression approach can be applied to these filtered waveforms to obtain an automatic, reliable and unbiased estimate of the peak latency and amplitude of the N1 wave, both in average and single-trial waveforms.
See also :
Hu L, Mouraux A, Hu Y, Iannetti GD. A novel approach for enhancing the signal-to-noise ratio and detecting automatically event-related potentials (ERPs) in single trials. Neuroimage (2010).
Flowchart describing the procedure developed to enhance the SNR of the N1 wave of LEPs (top panel), and to measure automatically its peak latency and amplitude in single trials (bottom panel).
Top panel. A time-frequency representation is obtained from group average waveform, using a continuous wavelet transform (CWT) (1), and a wavelet filter is generated by thresholding this time-frequency representation (2). This filter is applied (3) to the time-frequency representation obtained from single-trial LEP responses (4). Filtered single-trial LEP responses are then reconstructed in the time-domain using an inverse CWT (ICWT) (5), and finally averaged (6). This procedure generates both single-trial (5) and averaged (6) LEP responses with enhanced SNR.
Bottom panel: A multiple linear regression is applied to these single-trial LEPs with enhanced SNR to obtain a fast and unbiased estimate of the peak latency and amplitude of N1 waves. A regressor and its temporal derivative (0 to 0.25 s post-stimulus) are obtained from the across-trial average waveform (7). This basis set is then regressed (8) against the corresponding time window of each single LEP trial, thus yielding (9) a latency and amplitude value for each N1 peak.