Cerebrovascular Reactivity Mapping Using Resting-State fMRI in Early Chronic Traumatic Brain Injury
Poster No:
2337
Submission Type:
Abstract Submission
Authors:
Joon Yul Choi1, Savannah Steer2, Sung-min Kim1, Daniel Brennan2, Ines Guariguata2, Taehyun Kim1, Junghoon Kim2
Institutions:
1Yonsei University, Wonju, NA, 2The City University of New York, New York, NY
First Author:
Co-Author(s):
Introduction:
Altered cerebrovascular reactivity (CVR) may reflect microvascular injury after traumatic brain injury (TBI) and serve as an indicator of cerebrovascular health [1-4]. However, current gas inhalation CVR assessment protocols and paradigms pose barriers to using CVR in clinical settings; most require complex designs and training [5,6]. Recently, an alternative to gas inhalation-based CVR protocol was proposed using resting-state BOLD MRI [5,7]. This study aims to examine test/retest reliability of resting-state based CVR (RS-CVR) and its alterations in moderate-to-severe TBI (msTBI) patients compared to matched healthy controls (HC).
Methods:
Resting state fMRI (rsfMRI) images were acquired from 35 TBI patients at 3 months post-injury and 24 HC with the following parameters: TR/TE=3000/300 ms, matrix size=64x64, voxel size=3x3x3, and measurements=100. For 21 HC and 29 TBI subjects, two BOLD runs were available for reliability calculation. After preprocessing including motion correction and linear detrending, a bandpass filter with a range of 0 – 0.01164 Hz was applied to BOLD signals[7]. RS-CVR maps were obtained for each subject through a general linear model (GLM). To assess reliability, we computed intra-class correlations (ICCs) in two different ways: 1) ICC voxel-wise across runs within subjects and 2) ICC across subjects for each atlas based [8] region of interest (ROI). Aligned rank transformations (ART) ANOVA was conducted to compare differences in regional ICCs between groups and hemispheres. To compare RS-CVR values between TBI and HC, a GLM based on two sample t tests with covariates of age and gender was created. Whole-brain voxel-wise results from the GLM were evaluated using threshold free cluster enhancement (TFCE) inference and family-wise error (FWE) correction for multiple comparison correction.
Results:
The average ICC across voxels (calculated within subjects) was 0.75 ± 0.12 and 0.69 ± 0.15 in HC and TBI, respectively. Figure 1A shows the regional ICCs for the left and right hemisphere ROIs (ROI N=118) in HC and TBI. The ICCs for the left and right hemispheres in HC were 0.46 ± 0.16 and 0.48 ± 0.14, respectively while for TBI, these values were 0.37 ± 0.21 and 0.45 ± 0.17, respectively. The ART ANOVA revealed a significant difference in ICC between groups (F1 = 7.6, p < 0.01). However, in a subgroup analysis using only non-lesioned ROIs (N=50, Figure 1B), this group difference in ICC values disappeared, indicating that focal lesions lowered reliability of affected ROIs. When comparing RS-CVR between HC and TBI, HC exhibited significantly higher RS-CVR values in left cerebellum than TBI whereas TBI patients exhibited significantly higher RS-CVR values in cuneus than HC (Figure 2).
Conclusions:
Results from this study support previous studies' observations of good repeatability of within-subject RS-CVR maps quantified by voxelwise ICCs [5]. However, reliability measured by across-subject regional ICC values was substantially lower, especially for the ROIs affected by focal lesions. Group differences between msTBI and HC in the cerebellum in this study corroborate prior research implicating the cerebellum in TBI [4,9,10]. Lower RS-CVR values in the cuneus in HC compared to TBI might indicate that the cuneus is relatively preserved after msTBI.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s)
Modeling and Analysis Methods:
Task-Independent and Resting-State Analysis
Novel Imaging Acquisition Methods:
BOLD fMRI 1
Physiology, Metabolism and Neurotransmission :
Cerebral Metabolism and Hemodynamics 2
Keywords:
Cerebrovascular Disease
Degenerative Disease
FUNCTIONAL MRI
MRI
Trauma
1|2Indicates the priority used for review
Provide references using author date format
[2] Amyot F, Kenney K, Spessert E, et al., Assessment of cerebrovascular dysfunction after traumatic brain injury with fMRI and fNIRS, Neuroimage Clin, 25:102086, 2020.
[3] Chen JJ, Gauthier CJ, The role of cerebrovascular-reactivity mapping in functional MRI: Calibrated fMRI and resting-state fMRI, Front Physiol, 12:657362, 2021.
[4] Kenny K, Amyot F, Haber M, et al., Cerebral vascular injury in traumatic brain injury, Exp Neurol, 3:353-366, 2016.
[5] Liu P, Welch BG, Li Y, et al., Multiparametric imaging of brain hemodynamics and function using gas-inhalation MRI, Neuroimage, 146:715-723, 2017.
[6] Pinto J, Bright MG, Bulte DP, et al., Cerebrovascular reactivity mapping without gas challenges: A methodological guide, Front Physiol, 11:608475, 2021.
[7] Liu P, Liu G, Pinho MC, et al., Cerebrovascular reactivity mapping using resting-state BOLD functional MRI in healthy adults and patients with Moyamoya disease, Radiology, 299:419-425, 2021.
[8] Landman B, Warfiled S, MICCAI 2012 workshop on multi-atlas labeling, in: MICCAI Grand Challenge and Workshop on Multi-Atlas Labeling, CreateSpace Independent Publishing Platform, Nice, France, 2012.
[9] Gaggi NL, Ware JB, Dolui S, et al., Temporal dynamics of cerebral blood flow during the first year after moderate-severe traumatic brain injury: A longitudinal perfusion MRI study, Neuroimage Clin, 37:103344, 2023.
[10] Hillary FG, Roman CA, Venkatesan U, et al., Hyperconnectivity is a fundamental response to neurological disruption, Neuropsychology, 29:59-75, 2015.