Probabilistic DTI Tractography Shows Compromised Connectivity Between Regions of Atypically Increased Activity in Children with Autism Spectrum Disorder

Background:

Previous studies have shown white matter compromise in children and adults with autism spectrum disorder (ASD), which may relate to reduced connectivity and impaired function in distributed networks. However, tract-specific evidence remains limited in ASD.

Objectives:

To examine connectivity between regions with atypically increased activity during visual search in children with ASD, using probabilistic white matter fiber tracking.

Methods:

Diffusion tensor imaging (DTI) data of 17 children with ASD and 13 typically developing (TD) children were acquired from a 3T MRI scanner, using single-shot diffusion-weighted EPI pulse sequence with two degrees of diffusion weighting (b=0 and 2000 s/mm², 15 non-linear directions, four repetitions). Geometric distortions due to local magnetic field inhomogeneities were corrected using field maps.

Diffusion tensor tractography was performed using a probabilistic tracking approach from two regions of interest (ROIs) identified from a corresponding fMRI study, which showed atypically increased activation in ASD group in an fMRI study of visual search. Tracts were determined for pairs between a seed ROI in occipital white matter and four secondary ROIs in parietal and frontal lobes. Bayesian estimation of diffusion parameters using Markov Chain Monte Carlo sampling techniques and trilinear interpolation of the probability density functions were employed to determine the streamline between given ROIs. Volume and integrity of identified tracts were determined.

Results:

Significant group differences were detected for tract volume, fractional anisotropy (FA), mean diffusivity (MD), and radial diffusion. In the ASD group, FA was significantly lower (0.18+/-0.007 for ASD vs. 0.23+/-0.01 for TD, p=0.003); mean diffusivity was significantly higher (0.8+/-0.001 vs. 0.7+/-0.001 *10^-3 mm²/s, p=0.002), as was radial diffusion (0.7+/-0.001 vs. 0.6+/-0.001 *10^-3 mm²/s, p=0.008). Significantly decreased tract volume in ASD was also found (33078+/-16691 vs. 159237+/-36837 mm³, p=0.002).

Conclusions:

These results suggest that increased task-related regional activation may be accompanied by impaired network connectivity in children with ASD. Activation in atypical sites, as reported in previous fMRI studies of ASD, may therefore reflect isolated or ‘disconnected’ activity. Significantly higher radial diffusion in our ASD group may specifically suggest disruption of myelin sheaths responsible for maintaining axonal integrity.