Simultaneous cortical, subcortical, and brainstem mapping of sensory activation
Poster No:
2153
Submission Type:
Abstract Submission
Authors:
Neha Reddy1, Rebecca Clements2, Molly Bright1
Institutions:
1Northwestern University, Chicago, IL, 2Northwestern University, Evanston, IL
First Author:
Co-Author(s):
Introduction:
Whole-brain fMRI allows for complete mapping of neural systems that have key subcortical and brainstem components, such as sensorimotor pathways. However, subcortical and brainstem fMRI has historically been challenged by high physiological noise and small nuclei sizes, leading many studies to use a restricted field of view or move to 7T[2]. To address these challenges at 3T, we implemented a targeted scan protocol that provides a whole-brain field of view, sufficient in-plane spatial resolution in the brainstem, and multi-echo denoising for improved data quality. Here, we demonstrate our simultaneous cortical-subcortical-brainstem protocol to map sensory activation across the brain and test its ability to differentiate adjacent sensory nuclei in the brainstem: cuneate (upper extremity sensation) and gracile (lower extremity sensation).
Methods:
Data Collection: fMRI scans were collected in a Siemens 3T Prisma MRI system with a 32-channel head coil, using a multi-band multi-echo GRE EPI sequence: TR=2.2s, TEs=13.4/39.5/65.6ms, FA=90°, MB factor=2, voxel size=1.731x1.731x4mm3. Axial slices were aligned perpendicular to the base of the 4th ventricle. During each scan, a hand or foot was brushed at a rate of 1Hz: 20s on/20s off x 12 repeats. 10 healthy participants (4M,26±2y) underwent brushing of the right and left palm/fingers for 2 scans each, and 10 participants (5M,25±3y) underwent brushing of the right sole/toes for 2 scans. A T1-weighted structural image and field map were acquired to aid with registration and distortion correction.
Data Analysis[4,7]: The first 10 volumes of each fMRI scan were removed to allow for steady-state magnetization, then scans were distortion-corrected. Head-motion realignment parameters were computed for the first echo with reference to the initial Single Band reference image, then applied to all echoes. Optimally combined (OC) data were calculated, then converted to signal percentage change. Multi-echo ICA was performed (tedana[5]) and components were manually accepted or rejected as noise[6]. Subject-level activation was modeled with a sensory task regressor, motion parameters, Legendre polynomials up to 4th order, and rejected ICA components. Subject-level beta parameter and t-statistic maps were transformed to MNI space and averaged across sessions. Group-level activation was identified across the whole brain using AFNI 3dMEMA, and within a mask of the medulla using FSL randomise with threshold-free cluster enhancement.
Data Analysis[4,7]: The first 10 volumes of each fMRI scan were removed to allow for steady-state magnetization, then scans were distortion-corrected. Head-motion realignment parameters were computed for the first echo with reference to the initial Single Band reference image, then applied to all echoes. Optimally combined (OC) data were calculated, then converted to signal percentage change. Multi-echo ICA was performed (tedana[5]) and components were manually accepted or rejected as noise[6]. Subject-level activation was modeled with a sensory task regressor, motion parameters, Legendre polynomials up to 4th order, and rejected ICA components. Subject-level beta parameter and t-statistic maps were transformed to MNI space and averaged across sessions. Group-level activation was identified across the whole brain using AFNI 3dMEMA, and within a mask of the medulla using FSL randomise with threshold-free cluster enhancement.
Results:
For all stimuli, activation was detected in the sensorimotor cortices, putamen, and cerebellum. Thalamus activity was detected for hand stimuli (Fig1). With brainstem-specific analyses, activation was detected in the ipsilateral cuneate nuclei for hands and gracile nucleus for the foot; clusters did not overlap (Fig2).
Conclusions:
Cortical, subcortical, and brainstem activation findings for hand and foot stimuli aligned with expectations and previous findings[1]. The lack of thalamus findings for the right foot suggests that greater sample size may be needed, as foot sensory fibers are fewer than in the hand[3]. Sensory and motor systems are linked, demonstrated by activation detected in motor-related areas, such as the putamen and motor cortex, in addition to expected sensory areas. A similar single-echo acquisition protocol has been used previously to aid in full brain fMRI analyses[8]. We incorporated multi-echo denoising, shown to improve data quality[9], to enhance sensitivity to activation (particularly in the brainstem) with a small sample size.
Importantly, our protocol allowed for the lateralization of activity in the medulla for right vs left hand stimuli and, for the first time, differentiation between adjacent cuneate and gracile nuclei using fMRI. Our results demonstrate the feasibility of simultaneous whole-brain task-fMRI at 3T, with potential applications in investigating sensorimotor changes in clinical cohorts, such as stroke and Parkinson's, that have brainstem involvement.
Importantly, our protocol allowed for the lateralization of activity in the medulla for right vs left hand stimuli and, for the first time, differentiation between adjacent cuneate and gracile nuclei using fMRI. Our results demonstrate the feasibility of simultaneous whole-brain task-fMRI at 3T, with potential applications in investigating sensorimotor changes in clinical cohorts, such as stroke and Parkinson's, that have brainstem involvement.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 2
Methods Development
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Subcortical Structures 1
Novel Imaging Acquisition Methods:
BOLD fMRI
Perception, Attention and Motor Behavior:
Perception: Tactile/Somatosensory
Keywords:
Brainstem
Data analysis
FUNCTIONAL MRI
Somatosensory
Touch
1|2Indicates the priority used for review
Provide references using author date format
[2] Beissner F. Functional MRI of the Brainstem: Common Problems and their Solutions. Clin Neuroradiol 2015:251–7. https://doi.org/10.1007/s00062-015-0404-0.
[3] Corniani G. Tactile innervation densities across the whole body. J Neurophysiol 2020;124:1229–40. https://doi.org/10.1152/jn.00313.2020.
[4] Cox J.S. AFNI: Software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res 1996;29:162–73.
[5] Dupre E. TE-dependent analysis of multi-echo fMRI with tedana. J Open Source Softw 2021;6:1–5. https://doi.org/10.21105/joss.03103.
[6] Griffanti L. Hand classification of fMRI ICA noise components. Neuroimage 2017;154:188–205. https://doi.org/10.1016/j.neuroimage.2016.12.036.
[7] Jenkinson M. FSL. Neuroimage 2012;62:782–90. https://doi.org/10.1016/j.neuroimage.2011.09.015.
[8] Oliva V. Parallel cortical-brainstem pathways to attentional analgesia. Neuroimage 2021;226:117548. https://doi.org/10.1016/j.neuroimage.2020.117548.
[9] Reddy NA. Multi-echo ICA denoising increases tSNR in combined cortical and brainstem task-fMRI. Proc Annu Meet Organ Hum Brain Mapp 2023.