Single-subject Voxel-based Analysis for mTBI using Multi-shell Diffusion MRI
Stand-By Time
Wednesday, June 28, 2017: 12:45 PM - 2:45 PMSubmission No:
3266
Submission Type:
Abstract Submission
On Display:
Wednesday, June 28 & Thursday, June 29
Authors:
Xia Li1, Chitresh Bhushan1, Asha Singanamalli11, Ek Tan1, Jonathan Sperl2, Sumit Niogi3, John Tsiouris3, Pratik Mukherjee4, Joseph Masdeu5, Teena Shetty6, Luca Marinelli1
Institutions:
1GE Global Research, Niskayuna, United States, 2GE Global Research, Munich, Germany, 3Weill Cornell Medical Center, New York City, United States, 4University of California, San Francisco, San Francisco, United States, 5Houston Methodist, Houston, United States, 6Hospital for Special Surgery, New York City, United States
First Author:
Introduction:
Mild traumatic brain injury (mTBI) is a heterogeneous disease and it presents a significant challenge in the discovery of disease-sensitive imaging biomarkers. Multi-shell diffusion MRI (MSD-MRI) is a sensitive image acquisition technique with the ability to assess microstructural properties of white matter. In this study, we present a voxel-based approach using MSD data, which allows subject-specific and group-size analyses.
Methods:
MSD-MR data were acquired in 85 mTBI patients with no neurotrauma findings in structural MRI and 19 healthy controls at up to four encounters (around 3 days, 1 week, 3 weeks, and 3 months after injury) at three clinical sites using 3T GE MR750 clinical MR scanners (GE Healthcare, Milwaukee, WI). 140 diffusion weighted images along with eight b=0 images were acquired [1].
Diffusion metrics derived from MSD data included orthogonal kurtosis (KOrth), mean kurtosis (MZ), fractional anisotropy (FA), par-allel diffusivity (PD), and apparent diffusion coefficient (ADC). All the metrics were spatially normalized to the MNI space, with the ICBM white matter atlas [2], to obtain a voxel-by-voxel correspondence across subjects. For each voxel in white matter, we computed subject-wise z-score using the mean and standard deviation from the healthy control group. A z-score threshold of 2.56 was selected to define the outlier voxels for both the mTBI and control subjects. The percentage of outlier voxels in each ROI was then calculated and analyzed longitudinally.
Diffusion metrics derived from MSD data included orthogonal kurtosis (KOrth), mean kurtosis (MZ), fractional anisotropy (FA), par-allel diffusivity (PD), and apparent diffusion coefficient (ADC). All the metrics were spatially normalized to the MNI space, with the ICBM white matter atlas [2], to obtain a voxel-by-voxel correspondence across subjects. For each voxel in white matter, we computed subject-wise z-score using the mean and standard deviation from the healthy control group. A z-score threshold of 2.56 was selected to define the outlier voxels for both the mTBI and control subjects. The percentage of outlier voxels in each ROI was then calculated and analyzed longitudinally.
Results:
Fig 1 shows the analysis for three representative patients: KOrth, z-score map, and distribution of outlier percentages at four time points. The z-score maps display the locations and percentage of outlier voxels. Outlier percentage plots reflect the longitudinal change of metrics and highlighted the most affected white-matter regions. While the spatial distribution of outliers varies with sub-jects, we notice a trend towards certain white-matter structures being more commonly affected. Percentage of outlier voxels in each region for the control population is also shown for comparison.
Fig 2 displays the mean outlier percentage for all patients. The figure illustrates that some structures, such as posterior limb of internal capsule, superior corona radiata, cingulum bundle, and superior fronto-occipital fasciculus, have higher outlier fractions, compared to other regions, suggesting that these brain regions are commonly affected in the mTBI group. Moreover, a general trend is noticed from those regions: an increase in the outlier percentage acutely, followed by decrease from 7 to 90 days (E2 to E4). In a longitudinal anal-ysis with paired comparison of percentage outlier voxels across encounters, we found statistically significant difference between E1 and E2 for KOrth, MZ, and FA (p<0.05) with KOrth, MZ, and FA increasing from E1 to E2, and a trend of all metrics decreasing from E2 to E4. A possible interpretation of this result is temporary increased restriction of the extra-axonal water due to mild neuroinflam-mation and axonal swelling.
Fig 2 displays the mean outlier percentage for all patients. The figure illustrates that some structures, such as posterior limb of internal capsule, superior corona radiata, cingulum bundle, and superior fronto-occipital fasciculus, have higher outlier fractions, compared to other regions, suggesting that these brain regions are commonly affected in the mTBI group. Moreover, a general trend is noticed from those regions: an increase in the outlier percentage acutely, followed by decrease from 7 to 90 days (E2 to E4). In a longitudinal anal-ysis with paired comparison of percentage outlier voxels across encounters, we found statistically significant difference between E1 and E2 for KOrth, MZ, and FA (p<0.05) with KOrth, MZ, and FA increasing from E1 to E2, and a trend of all metrics decreasing from E2 to E4. A possible interpretation of this result is temporary increased restriction of the extra-axonal water due to mild neuroinflam-mation and axonal swelling.
·Fig 1. Korth, z-maps, and outlier distributions for 3 representative patients.
·Fig 2. Mean outlier distribution for all subjects. Notice that Korth emerges as a more sensitive measure of disease pathology than any of the DTI metrics.
Conclusions:
Microstructure-sensitive diffusion MRI is a promising approach to study mTBI. In single patients it allows for the identification of sub-regions differing from the healthy control group and therefore likely to be affected by mTBI. In addition, the group analysis suggests that few regions are commonly affected across patients and show similar longitudinal trend. This opens up possibility of identification of diagnostic and prognostic markers of mTBI that use both subject-specific and group-wise information and may ultimately help clinical management of individual patients.
Disorders of the Nervous System:
Traumatic Brain Injury 1
Imaging Methods:
Diffusion MRI
Modeling and Analysis Methods:
Methods Development 2
Keywords:
Data analysis
MRI
Trauma
White Matter
1|2Indicates the priority used for review
Would you accept an oral presentation if your abstract is selected for an oral session?
I would be willing to discuss my abstract with members of the press should my abstract be marked newsworthy:
Please indicate below if your study was a "resting state" or "task-activation” study.
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute the presentation in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels or other electronic media and on the OHBM website.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Internal Review Board (IRB) or Animal Use and Care Committee (AUCC) Approval. Please indicate approval below. Please note: Failure to have IRB or AUCC approval, if applicable will lead to automatic rejection of abstract.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references in author date format
2. http://www.loni.usc.edu/ICBM/Downloads/Downloads_DTI-81.shtml.