Visual mismatch negativity elicited by facial expressions: new evidence from the equiprobable paradigm

Converging evidence revealed that facial expressions are processed automatically. Recently, there is evidence that facial expressions might elicit the visual mismatch negativity (MMN), expression MMN (EMMN), reflecting that facial expression could be processed under non-attentional condition. In the present study, using a cross modality task we attempted to investigate whether there is a memory-comparison-based EMMN.

12 normal adults were instructed to simultaneously listen to a story and pay attention to a non-patterned white circle as a visual target interspersed among face stimuli. In the oddball block, the sad face was the deviant with a probability of 20% and the neutral face was the standard with a probability of 80%; in the control block, the identical sad face was presented with other four kinds of face stimuli with equal probability (20% for each). Electroencephalogram (EEG) was continuously recorded and ERPs (event-related potentials) in response to each kind of face stimuli were obtained. Oddball-EMMN in the oddball block was obtained by subtracting the ERPs elicited by the neutral faces (standard) from those by the sad faces (deviant), while controlled-EMMN was obtained by subtracting the ERPs elicited by the sad faces in the control block from those by the sad faces in the oddball block. Both EMMNs were measured and analyzed by ANOVAs (Analysis of Variance) with repeated measurements. sLORETA (standardized low-resolution brain electromagnetic tomography) was used to investigate the cortical generators of controlled-EMMN.

Both the oddball-EMMN in deviant-standard comparison and the controlled-EMMN in deviant-control comparison were observed at occipital-temporal regions with right hemisphere predominance. The oddball-EMMN was bigger and earlier than the controlled-EMMN because, besides the memory-based comparison, the former included a difference of refractoriness due to the distinction of presented probability between the deviant and standard face stimuli. The source analysis of controlled-EMMN indicated a current source primarily involved in posterior areas including superior temporal gyrus, postcentral gyrus, inferior parietal lobule as well as the insula.

The valid EMMN properly reflecting the memory-based comparison of facial expressions could be obtained, i.e., the controlled-EMMN.