Psychophysiology, 37:S17, 2000.
A direct comparison of Geodesic Sensor Net (128-channel) and conventional (30-channel) ERPs in tonal and phonetic oddball tasks
Jürgen Kayser, Craig E. Tenke, Nilabja Bhattacharya, Barbara K. Stuart, Jeffrey Hudson, Gerard E. Bruder
Department of Biopsychology, New York State Psychiatric Institute, New York, NY 10032, USA
Abstract
ERPs were recorded from 15 healthy right-handed adults during auditory oddball tasks in two sessions with: 1) a conventional 30-channel electrocap (NeuroScan); 2) a 128-channel net (Electrical Geodesics). Experimental features were held as constant as possible across systems. Stimuli were binaural complex tones or consonant-vowel syllables (960 trials, 20% targets per session). Response mode (left press, right press, silent count) was varied within participants, with the order counterbalanced across participants. Both data sets were referenced to nose tip and processed with the same PC-based software. The two systems revealed the same latencies for distinct ERP components (N1, N2, P3) for both group and individual averages. However, all components were reduced for the 128-channel system, with the extent of the reduction varying considerably across electrodes, conditions, and subjects. Covariance-based temporal PCAs (unscaled Varimax rotation) extracted essentially identical factors (N1, N2, early P3, late P3, slow wave) for both data sets. For the 30-channel system, ANOVA performed on factor scores confirmed task-related regional hemispheric asymmetries of N2 (tonal: frontotemporal R>L; phonetic: parietotemporal L>R) reported previously. For the 128-channel system, only the parietotemporal N2 asymmetry for syllables was found. Visual inspection of individual trial data suggests that the observed disparity reflects system differences in EEG stability at individual electrodes over time, resulting in system-dependent differences in the signal-to-noise ratio of ERP averages.
Keywords: dense electrode array; temporal PCA; oddball paradigm
