Aberrant somatosensory phase synchronization in children with hemiplegic cerebral palsy

Abstract

Children with hemiplegic cerebral palsy (HCP) often show disturbances of somatosensation. Despite extensive evidence of somatosensory deficits, neurophysiological alterations associated with somatosensory deficits in children with HCP have not been elucidated. Here, we aim to assess phase synchrony within and between contralateral primary (S1) and secondary (S2) somatosensory areas in children with HCP. Intra-regional and inter-regional phase synchronizations within and between S1 and S2 were estimated from somatosensory evoked fields (SEFs) in response to passive pneumatic stimulation of contralateral upper extremities and recorded with pediatric magnetoencephalography (MEG) in children with HCP and typically developing (TD) children. We found aberrant phase synchronizations within S1 and between S1 and S2 in both hemispheres in children with HCP. Specifically, the less-affected (LA) hemisphere demonstrated diminished phase synchronizations after the stimulus onset up to ~120 ms compared to the more-affected (MA) hemisphere and the dominant hemisphere of TD children, while the MA hemisphere showed enhanced phase synchronizations after ~100 ms compared to the LA hemisphere and the TD dominant hemisphere. Our findings indicate abnormal somatosensory functional connectivity in both hemispheres of children with HCP.

Keywords: Cerebral palsy; Functional connectivity; Magnetoencephalography; Phase synchronization; Primary somatosensory cortex; Secondary somatosensory cortex.

Figure 1:. Examples of SEFs and activations of contralateral S1 and S2 in a participant.

Average SEFs evoked by stimulation of a digit of a participant from all gradiometers (a) and the corresponding dSPM source localization of contralateral S1 and S2 (b). Arrows indicate peaks of cortical responses and dots the epicenter of the corresponding source activity (color coded). A 50% threshold was applied to the source dSPM estimations and only above-threshold activations were displayed.

Figure 2:. Group average of ITPC for the virtual channel placed in contralateral S1 and permutation t-test results.

(a) ITPC of the HCP LA S1; (b) ITPC of the HCP MA S1; (c) ITPC of the TD dominant S1; (d) statistical difference of ITPC between the HCP LA S1 and the HCP MA S1; (e) statistical difference of ITPC between the HCP LA S1 and the TD dominant S1; (f) statistical difference of ITPC between the HCP MA S1 and the TD dominant S1. Permutation t-values were thresholded by p< 0.05; t-values with corresponding p< 0.005 were highlighted with orange contours; 1,000 permutations were used in each test.

Figure 3:. Group average of PPC between virtual channels placed at contralateral S1 and S2 and permutation t-test results.

(a) PPC of the HCP LA hemisphere; (b) PPC of the HCP MA hemisphere; (c) PPC of the TD dominant hemisphere; (d) statistical difference of PPC between the HCP LA and MA hemispheres; (e) statistical difference of PPC between the HCP LA and the TD dominant hemispheres; (f) statistical difference of PPC between the HCP MA and the TD dominant hemispheres. Permutation t-values were thresholded by p< 0.05; t-values with corresponding p< 0.005 were highlighted with orange contours; 1,000 permutations were used in each test.