The effect of cerebrovascular reactivity on functional connectivity in children
Poster No:
2034
Submission Type:
Abstract Submission
Authors:
Quimby Lee1, Christine Nordahl2,3, Audrey Fan1,4
Institutions:
1Department of Neurology, University of California, Davis, Davis, CA, 2Department of Psychiatry and Behavioral Sciences, University of California, Davis, Davis, CA, 3MIND Institute, University of California, Davis, Davis, CA, 4Department of Biomedical Engineering, University of California, Davis, Davis, CA
First Author:
Co-Author(s):
Christine Nordahl
Department of Psychiatry and Behavioral Sciences, University of California, Davis|MIND Institute, University of California, Davis
Davis, CA|Davis, CA
Department of Psychiatry and Behavioral Sciences, University of California, Davis|MIND Institute, University of California, Davis
Davis, CA|Davis, CA
Audrey Fan, PhD
Department of Neurology, University of California, Davis|Department of Biomedical Engineering, University of California, Davis
Davis, CA|Davis, CA
Department of Neurology, University of California, Davis|Department of Biomedical Engineering, University of California, Davis
Davis, CA|Davis, CA
Introduction:
Functional connectivity (FC) based on resting-state functional magnetic resonance imaging (rsfMRI) is widely used to study neuronal network formation in typical development and neurodevelopmental conditions[1,2]. However, developmental heterogeneity may contribute to inconsistent FC findings in children. FC measures the temporal correlation between the blood oxygen level dependent (BOLD) signal of distinct brain regions. In fMRI, the BOLD signal indirectly measures neuronal activity by detecting changes in oxygenated blood flow, which occur through local blood vessel dilation. In addition to neuronal activity, vascular function can affect local oxygenated blood flow, and thus the regional BOLD signal. Cerebrovascular effects on FC have been identified in adults[3], but have yet to be examined in children to understand neuronal and vascular sources of connectivity in development, which has periods of rapid and heterogeneous brain vessel growth[4].
Cerebrovascular reactivity (CVR), or the capacity for blood vessels to dilate in response to a vasoactive stimulus, is a key indicator of brain vascular function. Previous fMRI studies have utilized the BOLD signal response to a vasodilation challenge, such as increased CO2 inhalation, to measure CVR in adults[5]. However, this CVR assessment is inaccessible to pediatric populations due to the challenges of respiratory monitoring and administration of air with elevated CO2 to children. To circumvent these challenges, recent studies propose using relative CVR (rCVR) mapping derived solely from resting-state BOLD fMRI signals[6,7]. With this innovation, CVR analyses can be conducted on standard rsfMRI scans, which facilitates CVR studies in populations with difficulty scanning (e.g. children), increases the utility of rsfMRI databases, and enables parallel analyses of CVR and FC within a single scan.
In this study, we leverage this resting-state CVR method and the Human Connectome Project-Development (HCP-D) to evaluate the effect of rCVR on FC in the default mode network (DMN) during middle childhood (9-12 years old). Characterizing cerebrovascular contributions to FC in pediatric populations will improve neuronal interpretations of FC during development and provide novel insight into neurovascular development.
Cerebrovascular reactivity (CVR), or the capacity for blood vessels to dilate in response to a vasoactive stimulus, is a key indicator of brain vascular function. Previous fMRI studies have utilized the BOLD signal response to a vasodilation challenge, such as increased CO2 inhalation, to measure CVR in adults[5]. However, this CVR assessment is inaccessible to pediatric populations due to the challenges of respiratory monitoring and administration of air with elevated CO2 to children. To circumvent these challenges, recent studies propose using relative CVR (rCVR) mapping derived solely from resting-state BOLD fMRI signals[6,7]. With this innovation, CVR analyses can be conducted on standard rsfMRI scans, which facilitates CVR studies in populations with difficulty scanning (e.g. children), increases the utility of rsfMRI databases, and enables parallel analyses of CVR and FC within a single scan.
In this study, we leverage this resting-state CVR method and the Human Connectome Project-Development (HCP-D) to evaluate the effect of rCVR on FC in the default mode network (DMN) during middle childhood (9-12 years old). Characterizing cerebrovascular contributions to FC in pediatric populations will improve neuronal interpretations of FC during development and provide novel insight into neurovascular development.
Methods:
HCP-D rsfMRI scans from children ages 9-12 years old were corrected for motion and slice-timing, and spatially smoothed (Gaussian kernel full-width-half-max=8mm). Volumes with framewise displacement (FD) greater than 0.2mm were scrubbed. After excluding scans with less than 75% of frames remaining after scrubbing, 83 participants (32 male) were evaluated.
To generate CVR maps from rsfMRI, we used voxel-wise general linear models that regress each voxel's BOLD signal on the frequency filtered (0.02-0.04Hz) global BOLD signal, which reflects natural changes in arterial CO2 during normal breathing8 (Fig1). CVR maps were normalized to global whole-brain CVR to produce rCVR maps. CVR calculation was performed in native fMRI space, then transformed into T1-anatomical and MNI standard space.
Before FC analysis, a high-pass frequency filter (0.008Hz) was applied to all voxels to eliminate linear trends in the temporal BOLD signal. FC was calculated as the z-scored Pearson correlation between the BOLD signal of each pair of regions in the DMN, identified by the Harvard-Oxford atlas. FC between each pair of regions was regressed on average rCVR in the two respective regions, age, sex, and mean FD.
To generate CVR maps from rsfMRI, we used voxel-wise general linear models that regress each voxel's BOLD signal on the frequency filtered (0.02-0.04Hz) global BOLD signal, which reflects natural changes in arterial CO2 during normal breathing8 (Fig1). CVR maps were normalized to global whole-brain CVR to produce rCVR maps. CVR calculation was performed in native fMRI space, then transformed into T1-anatomical and MNI standard space.
Before FC analysis, a high-pass frequency filter (0.008Hz) was applied to all voxels to eliminate linear trends in the temporal BOLD signal. FC was calculated as the z-scored Pearson correlation between the BOLD signal of each pair of regions in the DMN, identified by the Harvard-Oxford atlas. FC between each pair of regions was regressed on average rCVR in the two respective regions, age, sex, and mean FD.
Results:
There was a positive trend between rCVR and FC (Fig2) of the left angular gyrus (AG) and right AG (p=0.08) and of the left AG and the posterior cingulate (p=0.09). Effects of age, sex, and mean FD were not significant (p>0.1).
Conclusions:
In school-aged children, CVR may contribute to within-network FC between DMN regions, especially the left angular gyrus. Future studies will explore the effect of CVR on FC in different networks and at different stages of development, particularly at younger ages.
Lifespan Development:
Normal Brain Development: Fetus to Adolescence 2
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural)
Task-Independent and Resting-State Analysis 1
Novel Imaging Acquisition Methods:
BOLD fMRI
Physiology, Metabolism and Neurotransmission :
Cerebral Metabolism and Hemodynamics
Keywords:
Development
FUNCTIONAL MRI
PEDIATRIC
Other - Cerebrovascular Reactivity
1|2Indicates the priority used for review
Provide references using author date format
2. Hull JV, Dokovna LB, Jacokes ZJ, Torgerson CM, Irimia A, Horn JDV. (2017), 'Resting-State Functional Connectivity in Autism Spectrum Disorders: A Review,' Frontiers in Psychiatry, vol. 7, no. 205, PMCID: PMC5209637
3. Golestani AM, Kwinta JB, Strother SC, Khatamian YB, Chen JJ. (2016), 'The association between cerebrovascular reactivity and resting-state fMRI functional connectivity in healthy adults: The influence of basal carbon dioxide,' NeuroImage, vol. 132, pp. 301–313, PMCID: PMC5148617
4. Coelho-Santos V, Shih AY. (2020), 'Postnatal development of cerebrovascular structure and the neurogliovascular unit,' WIREs Developmental Biology, vol. 9, no. 2, pp. e363.
5. Liu P, De Vis JB, Lu H. (2019), 'Cerebrovascular reactivity (CVR) MRI with CO2 challenge: A technical review,' NeuroImage, vol.187, pp. 104–115, PMCID: PMC6150860
6. Liu P, Li Y, Pinho M, Park DC, Welch BG, Lu H. (2017), 'Cerebrovascular reactivity mapping without gas challenges,' NeuroImage, vol. 146, pp. 320–326, PMCID: PMC5321860
7. Liu P, Liu G, Pinho MC, Lin Z, Thomas BP, Rundle M, Park DC, Huang J, Welch BG, Lu H. (2021), 'Cerebrovascular Reactivity Mapping Using Resting-State BOLD Functional MRI in Healthy Adults and Patients with Moyamoya Disease,' Radiology, vol. 299, no. 2, pp. 419–425, PMCID: PMC8108558
8. Chang C, Glover GH. (2009), 'Relationship between respiration, end-tidal CO2, and BOLD signals in resting-state fMRI,' NeuroImage, vol. 47, no. 4, pp. 1381–1393, PMCID: PMC2721281