Pre-attentive Mismatch Response and Involuntary Attention Switching to a Deviance in an Earlier-Than-Usual Auditory Stimulus: An ERP Study

Abstract

An acoustic stimulus elicits an electroencephalographic response called auditory event-related potential (ERP). When some members of a stream of standard auditory stimuli are replaced randomly by a deviant stimulus and this stream is presented to a subject who ignores the stimuli, two different ERPs to deviant and standard stimuli are recorded. If the ERP to standard stimuli is subtracted from the ERP to deviant stimuli, the difference potential (DP) waveform typically exhibits a series of negative-positive-negative deflections called mismatch negativity (MMN), P3a, and reorienting negativity (RON), which are associated with pre-attentive change detection, involuntary attention switching, and reorienting of attention, respectively. The aim of the present study was to investigate how these pre-attentive processes are affected if the change occurs earlier than its usual timing implied by isochronous standard stimuli. In the MMN paradigm employed, 15% of the standards were randomly replaced by deviant stimuli which differed either in their pitch, their earlier onset time, or in both. Event-related responses to these three deviants [timely pitch change (R_TP), earlier onset (R_EO), earlier pitch change (R_EP)] and to standards (R_S) were recorded from 10 reading subjects. To maintain identical stimulation histories for the responses subtracted from each other, "deviant-standard" difference potentials (DP) for "timely" and "early" pitch deviances were derived as follows: DP_TP = R_TP - R_S and DP_EP = R_EP - R_EO. Interestingly, the MMN components of the DPs to timely and early pitch deviances had similar amplitudes, indicating that regularity of stimulus timing does not provide any benefit for the pre-attentive auditory change detection mechanism. However, different scalp current density (SCD) dynamics of the MMN/P3a complexes, elicited by timely and early pitch deviances, suggested that an auditory change in a stimulus occurring earlier-than-usual initiates a faster and more effective call-for-attention and causes stronger attention switching than a timely change. SCD results also indicated that the temporal, frontal, and parietal MMN components are simultaneously present rather than emerging sequentially in time, supporting the MMN models based on parallel deviance processing in the respective cortices. Similarity of the RONs to timely and early pitch deviances indicated that reorienting of attention is of the same strength in two cases.

Keywords: MMN; P3a; additivity; attention switching; inter-onset interval; pitch; regular stimulation.

Figure 1

Top panel: stimulation paradigm. Standard inter-onset interval (IOI) between successive tone-pips was 800 ms. Deviant stimuli which differed from standard (S) ones either only in their pitch (TP), or only in their earlier onset time (EO), or in both (EP) were randomly distributed in the sequence in such a way that the number of standards before a deviant changed randomly between 4 and 8, maintaining a probability of around 5% for each of the three deviants. Middle panel, left column: grand average event-related potentials (ERPs) to timely pitch-deviant, early onset timing-deviant, earlier pitch-deviant, and standard stimuli (R_TP, R_EO, R_EP, and R_S, respectively). Re-referenced to averaged earlobes. Middle panel, right column: the difference potentials (DPs) obtained for timely (IOI = 800 ms) and earlier (IOI = 500 ms) pitch deviances and for onset deviance due to shortening of the IOI, which were calculated as DP_TP = R_TP − R_S, DP_EP = R_EP − R_EO, and DP_EO = R_EO − R_S, respectively. Bottom panel: the two plots seen on the right of this panel with mismatch negativity (MMN), P3a, and reorienting negativity (RON) waves are the DPs calculated as DP_O&P (meas) = R_EP − R_S and DP_O&P (pred) = DP_EO + DP_TP, to obtain, respectively, the actually measured DP to “IOI&pitch” double-deviant, and its model waveform predicted by assuming additivity of the responses to single deviants (onset or pitch).

Figure 2

Scalp topographical plots of the grand average DPs to timely and earlier pitch deviances (DP_TP and DP_EP, respectively). Potentials are re-referenced to averaged earlobes. Global field power plots of the grand average DPs are at the bottom-left corner. At the top-right and bottom-right corners, scalp potential maps of the MMN components (mean over 120–180 ms latency range) in the two DPs are given.

Figure 3

Laplacian maps displaying the scalp current density (SCD) distributions of the Grand average MMN and P3a components in the ERPs to timely and earlier pitch deviances (TP and EP, respectively). Maps were obtained for the mean Laplacians over the time intervals indicated by gray stripes. Please note that the current density color scale is the same for all maps. N1 deflection of the ERP to standard stimulus is given in broken lines for comparison of its latency with the latency of MMN.