Cortical Generators and Connections Underlying Phoneme Perception: A Mismatch Negativity and P300 Investigation

Yana Criel,Emma Depuydt,Marijke Miatton,Patrick Santens,Pieter van Mierlo,Miet De Letter
DOI: https://doi.org/10.1007/s10548-024-01065-z
Abstract:The cortical generators of the pure tone MMN and P300 have been thoroughly studied. Their nature and interaction with respect to phoneme perception, however, is poorly understood. Accordingly, the cortical sources and functional connections that underlie the MMN and P300 in relation to passive and active speech sound perception were identified. An inattentive and attentive phonemic oddball paradigm, eliciting a MMN and P300 respectively, were administered in 60 healthy adults during simultaneous high-density EEG recording. For both the MMN and P300, eLORETA source reconstruction was performed. The maximal cross-correlation was calculated between ROI-pairs to investigate inter-regional functional connectivity specific to passive and active deviant processing. MMN activation clusters were identified in the temporal (insula, superior temporal gyrus and temporal pole), frontal (rostral middle frontal and pars opercularis) and parietal (postcentral and supramarginal gyrus) cortex. Passive discrimination of deviant phonemes was aided by a network connecting right temporoparietal cortices to left frontal areas. For the P300, clusters with significantly higher activity were found in the frontal (caudal middle frontal and precentral), parietal (precuneus) and cingulate (posterior and isthmus) cortex. Significant intra- and interhemispheric connections between parietal, cingulate and occipital regions constituted the network governing active phonemic target detection. A predominantly bilateral network was found to underly both the MMN and P300. While passive phoneme discrimination is aided by a fronto-temporo-parietal network, active categorization calls on a network entailing fronto-parieto-cingulate cortices. Neural processing of phonemic contrasts, as reflected by the MMN and P300, does not appear to show pronounced lateralization to the language-dominant hemisphere.
What problem does this paper attempt to address?