Program No. 875.17. 2005 Abstract Viewer and Itinerary Planner. Washington, DC: Society for Neuroscience. CD-ROM.
Separable parietal and temporal generators of P3 in dichotic tonal and phonetic oddball tasks: A Principal Components Analysis (PCA) of Current Source Density (CSD) waveforms in healthy adults and depressed patients
C.E. Tenke1,2, J. Kayser1,2, S.A. Shankman1, C.B. Griggs1, P. Leite1 G.E. Bruder1,2
1Department of Biopsychology, New York State Psychiatric Institute, NY, NY, USA
2Department of Psychiatry, Columbia University College of Physicians and Surgeons, NY, NY, USA
Abstract
ERPs (31-channel; 200 sps; blink correction) were recorded from 26 healthy adults and 38 depressed outpatients in an oddball paradigm that exploits the perceptual challenge of dichotic stimulation. Stimuli consisted of complex tones or consonant-vowel syllables in separate blocks. For the tonal task, nontargets were tone pairs (G and B above middle C), presented simultaneously to each ear (L/R) in an alternating series (L-G/R-B or L-B/R-G), with a fixed 2-s SOA. A target tone (A) replaced one of the pair (e.g., L-A/R-B) on 20% of the trials. Subjects responded with a button press to targets (hand counterbalanced within subjects). For the phonetic task, spoken phonemes were edited to match the duration and loudness of the tones. Paired nontargets were unvoiced (/ba/, /da/), and the target syllable was voiced (/ta/). Total correct performance was sufficient to record ERPs (tonal vs. phonetic, controls: 95% vs. 98%, patient: 93% vs. 96%). Reference-free CSDs were simplified and measured using unrestricted,covariance-based PCA with Varimax rotation. Target-related N2/P3 was represented by a lowvariance CSD factor peaking at 245 ms (F245; 3.5% variance; anterior sink; task-specific topography), followed by a prominent CSD factor peaking at 440 ms (F440; 30.2%; Pz maximum). Both factors were reduced in patients compared with controls. Sharply contrasting with binaural oddball findings, F440 included a secondary, left-lateralized temporal lobe topography, that was prominent for both targets and nontargets. Group differences in F440 were preserved when performance was included in the design (i.e., good performers >= 98% for tones), but eliminated for F245. Results implicate a distinct temporal generator of scalp-recorded P3 in a dichotic oddball task, and are consistent with prior findings of reduced P3 amplitude in depression.
Support Contributed By: MH36295 and MH50715
Keywords: P300, surface Laplacian, Event Related Potentials (ERP), dichotic listening