Memory-based mismatch response to changes in duration of auditory stimuli: an MEG study

Objective: Differences in physical features and occurrence probability between standards and deviants in oddball paradigms provide contributions to magnetic mismatch negativity (MMNm). We aimed to reduce these influential factors and extract memory-based MMNm by adding a control paradigm. Methods: Magnetoencephalographic responses were recorded in 13 healthy adults with an oddball paradigm (125-ms standard and 50-ms deviant tones) and an equiprobable control paradigm (50-ms control and four other duration-varying tones). The stimulus onset asynchrony was 500 ms. Controlled MMNm was obtained by subtracting control-evoked responses from deviant-evoked responses. Results: With respect to the onset of stimulus difference, the peak latency of controlled MMNm was compatible with previous intracranial MMN recordings. Both controlled and traditional MMNm were generated around the superior temporal cortex, whereas the controlled MMNm amplitude was about 70% of traditional MMNm amplitude. Right-hemispheric dominance was observed in traditional MMNm but not in controlled MMNm. N100m amplitude was smaller in standard-evoked than in deviant- or control-evoked responses. Conclusions: Controlled MMNm reflects memory-based processing of duration changes, whereas traditional MMNm additionally involves non-memory activations related to differential refractoriness states and physical properties between standard and deviant stimuli. Significance: The memory-based processing of auditory deviants may be preferentially extracted by adding a control paradigm.