3419

Brain White Matter Abnormality Induced by Chronic Spinal Cord Injury in the Pediatric Population: A Tract Based Spatial Statistic Study
Joshua Fisher1, Mahdi Alizadeh1, Devon Middleton1, Caio Matias1, MJ Mulcahey1, Feroze Mohamed1, and Laura Krisa1

1Thomas Jefferson University, Philadelphia, PA, United States

Synopsis

Few studies have quantified the change in cortical white matter tracts following chronic spinal cord injury in a pediatric population. Additionally, no work has been done to compare chronic SCI subjects with different American Spinal Injury Association Impairment (AIS) scale classifications. We hypothesized that these cortical changes can be detected using tract-based spatial statistics (TBSS). Our efforts revealed that significant changes in fractional anisotropy occur in several motor and sensory related regions. We conclude that TBSS can be effectively used to identify alterations in brain microstructure in a chronic pediatric spinal cord injury population.

Objective

Diffusion tensor imaging (DTI) is widely used in neurological and neuropsychiatric diseases, such as Parkinson’s disease (PD), epilepsy and traumatic brain injury (TBI) [1]. Previous studies have shown DTI as a quantitative biomarker to measure both axonal and myelin integrity of the brain following spinal cord injury (SCI) [2]. Tract based spatial statistics (TBSS) is a DTI based processing technique that aims to improve the objectivity and interpretability of analysis of multi-subject diffusion imaging studies. To the best of our knowledge, no study has investigated tract based analysis of the brain in pediatric SCI subjects. In this study, we used TBSS to detect cortical white matter changes following SCI.

Design/Methods

Eighteen subjects with chronic SCI aged 8-20 years old (mean16.5 years) were scanned using a DTI protocol in a 3.0T Siemens Verio MR scanner. All subjects underwent a complete International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) examination to determine the level and severity of injury. Five subjects were classified as American Spinal Injury Association Impairment Scale (AIS) A, nine AIS B, and four as AISC/D. This study was approved by the university’s IRB and the appropriate consent/assent was obtained. Imaging parameters used for data collection were as followed: 20 directions, b=1000s/mm2, voxel =1.8mm x1.8mm, slice = 5mm, TE = 95 ms, TR = 4300 ms, slices = 30,TA = 4:45. The raw data set of the diffusion volumes were first corrected for eddy current distortions and motion artifacts. The diffusion tensor of each voxel was then fit, and the fractional anisotropy (FA) map was calculated based on the eigenvalues of diffusion tensors. To generate TBSS, FA map were aligned to MNI space [2,3] and a skeleton FA mask based on the mean FA images for all subjects was generated to serve as a white matter mask for the group [3]. Nonparametric permutation tests were used for voxel-wise statistical analysis of the FA skeletons between the different groups. In addition, a two-tailed t-test was applied to extract voxels with significant differences between groups. The significance threshold for group differences (AIS A, B and C/D) was set at p < 0.05.

Results

Significant changes in FA are identified in the frontal lobe and in each of the three convexity frontal gyri (i.e., superior, middle, and inferior). Altered regions in the temporal, occipital and parietal lobe were also identified. Overall, significant changes were identified in all three group comparisons A versus AIS C/D, AIS A vs AIS B and AIS B vs AIS C/D. Also, these changes have been shown in both hemispheres however, changes are greater in one hemisphere compared to the other.

Conclusion

These results suggest that cortical white matter is altered differently between people with different AIS classifications. While studying all brain networks is not feasible, TBSS has potential to serve as a screening tool to identify regions of interest. Further studies directed at investigating these regions may provide new insights to understanding white matter abnormality as a function of injury severity. While these preliminary results are encouraging, they warrant further studies with a larger population.

Learning Objective

Learning Objective Discuss how the severity of spinal cord injury can lead to white matter changes in the cerebral cortex using TBSS.

Keywords: Pediatric, diffusion tensor imaging, Tract Based Spatial Statistic

Acknowledgements

Support: The study was funded by the Shriners Hospitals for Children Research Grant 8509

References

[1] Charlotte L. Rae, Marta M. Correia, Ellemarije Altena, Laura E. Hughes, Roger A. Barker, James B. Rowe. White matter pathology in Parkinson's disease: The effect of imaging protocol differences and relevance to executive function, Neuroimage, 2012; 62(3-2): 1675–1684.

[2] Niels K. Focke, Mahinda Yogarajah, Silvia B. Bonelli, Philippa A. Bartlett, Mark R. Symms, John S. Duncan. Voxel-based diffusion tensor imaging in patients with mesial temporal lobe epilepsy and hippocampal sclerosis, NeuroImage, 2008; 40 (2): 728-737.

[3] Peng Sun, Rory K. J. Murphy, Paul Gamble, Ajit George, Sheng-Kwei Song and Wilson Z. Ray. Diffusion Assessment of Cortical Changes, Induced by Traumatic Spinal Cord Injury, brain sciences, 2017;7(21):1-13

Figures

Table 1. TBSS analysis results between SCI patients with AIS A vs. B, AIS A vs. C/D and AIS B vs. C/D represented with the number of voxels in each area (Vx). Voxel numbers represents the significant difference in FA. P <0.05.

Figure 1: Areas of statistical change identified by TBSS. Coronal images of areas of statistical change in AIS A versus AIS C/D subjects. Images are aligned from posterior (A) to anterior (F). Increases in FA are shown in red. Decreases in FA are shown in blue. Green represents the mean FA skeleton. Slice number is shown in bottom right.

Figure 2: Areas of statistical change identified by TBSS. Coronal images of areas of statistical change in AIS A versus AIS B subjects. Images are aligned from posterior (A) to anterior (F). Increases in FA are shown in red. Decreases in FA are shown in blue. Green represents the mean FA skeleton. Slice number is shown in bottom right.

Figure 3: Areas of statistical change identified by TBSS. Coronal images of areas of statistical change in AIS B versus AIS C/D subjects. Images are aligned from posterior (A) to anterior (F). Increases in FA are shown in red. Decreases in FA are shown in blue. Green represents the mean FA skeleton. Slice number is shown in bottom right.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
3419