Role of secondary somatosensory cortex in haptic change detection: a MEG study

• ¹ Université libre de Bruxelles (ULB), UNI – ULB Neuroscience Institute, Belgium

Introduction Evolving in an environment requests crucial aptitude to detect novelty. The mismatch negativity (MMN) is an automatic brain response elicited by a novel sensory stimulation in a familiar environment and represents a neurophysiological marker of change-detection. Firstly discovered in the auditory modality, less is known about somatosensory MMN (sMMN). Here, we studied using magnetoencephalography (MEG) the spatio-temporal dynamics and the neural mechanisms at the basis of sMMN. Methods Somatosensory evoked magnetic fields (SEF) were recorded in 16 healthy right-handed (aged 29 +/- 3 years, 7 females) using a whole-scalp MEG (bandpass: 0.1-330 Hz, sampling rate: 1 kHz; Eletka) while they underwent repetitive unilateral tactile stimulations. Standard stimulations (standards) corresponded to a pneumatic tactile stimulation (covering area: 1 cm2, intensity: 2 bars, duration: 50 ms) applied to the right index fingertip. Deviant stimulations (deviants) corresponded to similar stimulation but simultaneously applied to the right forefinger fingertip and middle phalanx. SEFs were investigated in 3 randomized conditions: (1) Oddball: 100 blocks of four standards followed by a deviant (inter-stimulus interval (ISI): 0.5 sec, inter-block interval (IBI): 0.8 sec), (2) Dual: 80 blocks comprising one standard immediately followed by a deviant (ISI: 0.5 sec, IBI: 1-6 sec), and (3) Alone: 80 deviants alone (ISI: 1-6 sec). Furthermore, 10 subjects (aged 29.6 +/- 4 years, 3 females) underwent a complementary oddball condition (Long IBI Oddball) that was similar to Oddball except that IBI was longer and randomly set between 1.5 and 2.5 s. MEG signals were preprocessed and analyzed using conventional methods of signal processing used to evidence MMN. Differences between standards or deviants obtained in the different conditions were assessed at the sensor and individual levels using non-parametric clustering tests. Generators of the magnetic sMMN (msMMN) were localized using classical equivalent current dipole (ECD) modeling. Source strength waveforms of corresponding ECDs were then computed for each subject and condition over the considered epochs. Paired t-tests were used to search for significant differences in sources intensity. Results were considered significant at p<0.05. Results Figures 1 and 2 summarize the results obtained in this study. In Oddball, a significant msMMN peaking around 60-180 ms post-deviant onset was consistently observed across subjects with a msMMN neural generator located in the secondary somatosensory cortex (SII) contralateral to the stimulation. The intensity of the SII source generating the msMMN was significantly higher for deviants than standards. In Long IBI Oddball, similar results were observed but only when comparing the fourth standards with the deviants. Interestingly, comparisons between SEFs obtain in each condition demonstrated that a major response suppression phenomenon took place in SII cortex from the second tactile stimulus even when the tactile stimulus presented shortly after a first standard was a deviant (Dual). This suppression phenomenon was not observed in Oddball where there was no difference between the first and the fourth standards. Conclusion This study demonstrates that tactile deviants during a unilateral oddball paradigm generate a msMMN peaking at 60-180 ms post-deviant onset with SII cortex is the main msMMN neural generator. Results therefore suggest that SII cortex plays a key role for tactile change detection in humans under the framework of predictive coding. Figure legend: Figure 1 sMMN results obtained for a typical subject in Oddball Sensor space: 1. Spatial distribution of SEFs for standards and deviants, and the computed sMMN amongst the 306 MEG sensors. 2. Enlarged planar gradiometer showing the difference between standards and deviants, as well as the computed sMMN. 3. Location of the significant cluster (difference between standards and deviants) over the left fronto-parieto-temporal sensors. Source space: 4. ECD corresponding to the neural generator of the sMMN (left hemisphere sagittal slice). 6. Comparison between SII source waveforms of standards and deviants. The blue line indicates timing of the significant cluster in the sensor space. Figure 2 This figure summarizes the results of this study. SEFs elicited by standards are represented in blue while those elicited by deviants are represented in red.