Atypical Brain Response to Simultaneous Touch and Sound In Children with Sensory Processing Differences: A Multisensory Integration Functional Imaging Study

Background:  Atypical auditory and tactile response behaviors are ubiquitous in autism spectrum disorders (ASD).  These behaviors are thought to reflect early unimodal and multimodal processing differences that may underlie the core ASD deficits in social skills, communication and restricted interests/behaviors.  In fact, over 30 years of evoked potential research suggests cortical differences in the processing of sound, sight, and touch for children with ASD.  A recent ASD study by Russo et. al (2010) suggests delayed multisensory integration.  We have previously reported decreased functional brain connectivity as well as decreased early auditory cortex evoked response amplitudes in children with sensory processing differences (SPD).  However, multisensory integration of auditory and tactile information has not previously been explored in this population. Children with SPD, who do not meet ASD criteria, provide a unique opportunity for probing the neural underpinnings of abnormal sensory processing behaviors and can help to understanding atypical sensory behaviors, measure treatment response, and shape more functional neural pathways in susceptible individuals

Objectives:  Our goal is to investigate whether children with SPD show neural differences in multisensory integration.  We used magnetoencephalographic imaging (MEG-I), a technique that examines changes in cortical activity on a millisecond level, to create maps of cortical activation for children with SPD and matched controls during unimodal and bimodal auditory and tactile stimulation We hypothesized that children with SPD would exhibit reduced activation in the oscillatory frequencies thought to underlie long-range cortical communication (gamma band) in multimodal regions.

Methods:  Brain responses to simultaneous monaurally tone (-45dB) and finger tap (17PSI) were recorded for the SPD group (n=10, mean age=10.6 years) and the HC group (n=10, mean age=10.0 years) using a 275-channel whole-head biomagnetometer system at a sampling rate of 1200Hz. Epochs of 900ms (400ms pre-stimulus) were collected and filtered 2-40Hz. For evoked field analysis, MEG sensor data was averaged for each subject and the latency and amplitude(root mean squared; RMS) of the M110, M164, and M190 peaks at the left supramarginal gyrus were compared using a mixed effects model. MEG data was also reconstructed in the time-frequency domain using an adaptive spatial filtering technique, which allowed us to observe induced (non phase-locked) responses, measured as change in the modulation of oscillatory activity. We investigated high gamma (65-115Hz) and ultra-high gamma (115-150Hz) frequency bins. Time-frequency reconstructions were spatially normalized and entered into a group analysis using statistical non-parametric mapping. 

Results:  The SP group showed trends toward reduced M110 amplitudes with increased later activity at M190 within the left supramarginal gyrus, a region previously identified as underconnected in the SPD group (p=.06).  Decreased high-frequency oscillations were also identified in the SPD group over medial posterior parietal cortex (p=0.01). 

Conclusions:  This study suggests that children with SDP, like children with ASD, have measurable and reduced early MSI of simple auditory and tactile stimuli.  The time frequency analysis made possible by MEG-I will direct our ongoing investigation of how brain regions differ in their processing of sensory information both in time as well as in oscillatory frequency.