Preattentive auditory processing – a comparison between traditional and optimized paradigms in EEG and MEG

Aims: The Mismatch negativity (MMN) is an event-related potential (ERP) which provides an index of automatic context-dependent information processing and auditory sensory memory. MMN studies using traditional „oddball“ paradigms (e.g. 80% standards, 20% deviants) replicated deficits in preattentive information processing especially in schizophrenia. In a so-called „optimum“ paradigm with only 50% standards but 50% deviants (Näätänen et al. 2003) MMN responses are still maintained but assessment time is considerably shortened with advantage of investigating several deviants simultaneously. The present study compared the „oddball“ with the „optimum“ paradigm as well as EEG and MEG measures. We hypothesized the highest sensitivity in the „optimum“ paradigm with MEG, i.e. the best signal-to-noise ratio.

Methods: 12 healthy right-handed volunteers were investigated in a combined EEG und MEG study comparing between (a) the traditional „oddball“ paradigm and the „optimum“ paradigm, (b) between several deviants and (c) between the methods EEG (64-channel) and 148-channel whole-head MEG. Data analysis was preformed with Matlab (Mathworks Inc. USA).

Results: Both, the evoked potentials/fields for EEG/MEG showed the typical topographic distribution for auditory ERP i.e. MMN in the preliminary data analysis. Prominent peaks of the averaged dipole strengths corresponded to P50 and N100 components (peak latency: 50–80 and 100–130 ms after stimulus onset). MEG responses showed relatively stronger and prolonged P50 responses as compared to the ERPs suggesting a higher sensitivity to early cortical processes. In MEG deviants elicited reliable mismatch responses over both hemispheres, but presenting a stronger activation for the right auditory area (max Z value traditional: 3.8; optimized: 5.6).

Conclusion: This study employed a novel design in a combined EEG & MEG setting to compare the efficiency of „oddball“ and „optimum“ paradigm. MEG may reflect earlier processes than EEG. The larger differences in latency may suggest that the deviance detection as measured by the traditional oddball design reflect more specific stimulus representations. Direct statistical comparison and source analysis should verify these suggestions and allow reliable clinical applications.

Supported by IZKF BIOMAT (VV N68h), AstraZeneca