Sensitivity of EEG and MEG to the N1 and P2 auditory evoked responses modulated by spectral complexity of sounds

Abstract

Acoustic complexity of a stimulus has been shown to modulate the electromagnetic N1 (latency approximately 110 ms) and P2 (latency 190 ms) auditory evoked responses. We compared the relative sensitivity of electroencephalography (EEG) and magnetoencephalography (MEG) to these neural correlates of sensation. Simultaneous EEG and MEG were recorded while participants listened to three variants of a piano tone. The piano stimuli differed in their number of harmonics: the fundamental frequency (f ( 0 )), only, or f ( 0 ) and the first two or eight harmonics. The root mean square (RMS) of the amplitude of P2 but not N1 increased with spectral complexity of the piano tones in EEG and MEG. The RMS increase for P2 was more prominent in EEG than MEG, suggesting important radial sources contributing to the P2 only in EEG. Source analysis revealing contributions from radial and tangential sources was conducted to test this hypothesis. Source waveforms revealed a significant increase in the P2 radial source amplitude in EEG with increased spectral complexity of piano tones. The P2 of the tangential source waveforms also increased in amplitude with increased spectral complexity in EEG and MEG. The P2 auditory evoked response is thus represented by both tangential (gyri) and radial (sulci) activities. The radial contribution is expressed preferentially in EEG, highlighting the importance of combining EEG with MEG where complex source configurations are suspected.

Fig. 1

(A) Depiction of regional sources. Each regional source consists of two or three dipoles, one radial (R, EEG only) and two tangential (T1 and T2, EEG & MEG). (B) Example of auditory evoked potential (overlay of 32 traces) and field (overlay of 151 traces) for one subject, depicting the N1/P2 and N1m/P2m, respectively. Dotted line depicts the sound onset

Fig. 2

(A) Source locations for the N1, N1m, P2, P2m auditory evoked responses. Dotted lines are drawn for symmetry reference. (B) Between subject mean standard deviation values (collapsing over tones and hemispheres) for source locations in the x (medial–lateral), y (anterior–posterior) and z (inferior–posterior). m (as in Xm) denotes the magnetic value

Fig. 3

(A) Grand mean (n = 16) of N1 (top) and P2 (bottom) RMS values and error bars (one standard error) for piano0, piano2 and piano tones in EEG and MEG. (B) The grand mean (n = 16) of the N1 (top) and P2 (bottom) dipole moments for the radial (EEG only) and tangential components and error bars (one standard error) for piano0, piano2 and piano tones in EEG and MEG