Direct imaging of macrovascular and microvascular contributions to BOLD fMRI in layers IV–V of the rat whisker–barrel cortex
Highlights
► Map individual vessel fMRI responses in deep cortical layers. ► Penetrating macro vessels dominate the deep-layer peak BOLD responses. ► Decouple the time-dependent macro/micro-vascular contribution to BOLD signals. ► BOLD signal at 0.8 s after stimulation is less contributed by macrovasculature.
Introduction
BOLD-fMRI has become one of the most common techniques to map brain function (Bandettini et al., 1992, Kwong et al., 1992, Ogawa et al., 1992). The BOLD contrast relies on detecting the hemodynamic response to changes in neural activity. Upon functional activation, the evoked neuronal/metabolic response increases the local release of vasoactive agents, causing vasodilation and local increases in cerebral blood flow (CBF) (Attwell and Iadecola, 2002, Attwell et al., 2010, Buxton, 2002, van Zijl et al., 1998). The increased inflow of oxygenated arterial blood leads to an increase in the oxy-to-deoxyhemoglobin ratio (HbO/Hb) along the local vasculature (Fox and Raichle, 1986, Malonek et al., 1997), producing a signal increase in T2⁎-weighted MR images (Ogawa et al., 1990, Thulborn et al., 1982). Thus, the spatial specificity of fMRI is limited by the spatiotemporal dynamics of the functional hemodynamic response (HRF) (Harel et al., 2006, Kim and Ugurbil, 2003, Ugurbil et al., 2003).
Previous measurements of the functional hemodynamic response to visual stimulation estimated the full-width-at-half-maximum (FWHM) of the BOLD spatial point-spread function (PSF) to be in the range of 1.7–3.9 mm in human subjects (Engel et al., 1997, Kim et al., 2004, Parkes et al., 2005, Shmuel et al., 2007, Turner, 2002, Yacoub et al., 2005) and 470 μm in the cat visual cortex (Duong et al., 2001). Recently, in studying the reorganization of the primary somatosensory cortex a PSF of approximately 300–400 μm FWHM was also reported (Yu et al., 2010). Thus, it is clear that the signal spread through the vasculature can limit the spatial specificity of BOLD fMRI. In particular, large draining veins on the surface lead to mislocalization of BOLD signals even at high magnetic field (Keilholz et al., 2006, Kim et al., 1994, Kim et al., 2004, Lai et al., 1993, Lu et al., 2004, Ugurbil et al., 2003). In contrast to BOLD fMRI, other methods, such as arterial spin labeling MRI or cerebral blood volume MRI, have been proposed to suppress the macrovascular effect (Bolan et al., 2006, Duong et al., 2001, Lu et al., 2003, Lu et al., 2004, Williams et al., 1992, Zhao et al., 2005). However, BOLD fMRI is still the most popular functional mapping method due to the robust signal and ease of acquisition. Two major strategies have been used to eliminate the contribution of large draining veins. One strategy relies on differentiating the intra/extra-vascular dephasing properties of spins surrounding large vessels, such as spin-echo sequences or phase-dependent elimination of voxels (Duong et al., 2003, Goense and Logothetis, 2006, Menon, 2002, Ugurbil et al., 2003, Zhao et al., 2004). Another strategy relies on measuring the spatial and temporal response of the BOLD HRF. By estimating the time-dependent PSF, the initial phase of the BOLD HRF has been reported to be more spatially specific than the later phase (Goodyear and Menon, 2001, Lee et al., 1995, Shmuel et al., 2007, Silva and Koretsky, 2002). However, due to the limited spatial resolution of typical fMRI experiments, there are no studies to analyze the time-dependent contributions from microvasculature and macrovasculature to BOLD signals in the deep cortex at the level of individual vessels with BOLD fMRI.
Previously, it has been reported that there is an early onset of positive BOLD signal changes in the deep somatosensory cortex of the rat ~ 600 ms after stimulation (Hirano et al., 2011, Silva and Koretsky, 2002). This early positive BOLD response is significantly shorter than the half-transit time of ~ 1.7 s from arteries to veins measured by in vivo optical imaging (Hutchinson et al., 2006, Masamoto et al., 2010), providing evidence that fMRI changes occur before the oxy-hemoglobin can enter large veins. Based on these previous results, it may be expected that analysis of the BOLD HRF in the interval of 0.6–1.7 s following stimulus onset will reduce the macrovascular contributions to the BOLD signal. In the present work, fMRI experiments at high spatial (150 x 150 × 500 μm) and temporal (200 ms) resolution were performed to investigate the spatiotemporal characteristics of the BOLD HRF in the time interval of 0.6–1.6 s following stimulation of the whisker pad in chloralose-anesthetized rats.
Section snippets
Animal usage
Fifteen male Sprague–Dawley rats were imaged at 8–9 weeks of age. The high spatiotemporal EPI images (200 ms TR) were acquired from eleven rats. Coronal EPI slices were acquired from 7 rats and horizontal EPI slices were acquired from 6 rats (Two among the eleven rats were imaged at both coronal and horizontal orientation). The EPI images (800 ms TR) were acquired from the other 4 rats.
Animal preparation for functional MRI
A detailed procedure is described in a previous study (Yu et al., 2010). To briefly describe the preparation
The spatiotemporal pattern of hemodynamic responses of the barrel cortex
High-resolution coronal EPI images (150 × 150 × 500 μm) were acquired to visualize intracortical vessels penetrating the barrel cortex (Fig. 1A). The dark stripes in the barrel cortex were composed of voxels with lower signal intensity due to faster T2* decay. For visualization purposes, signal intensity correction was performed to remove the heterogeneous B1 signal profile of the surface coil (Fig. 1B). This procedure allowed better visualization of the intracortical macro vessels. The spatial
Discussion
The present study describes the spatiotemporal dynamics of the neurovascular coupling in the deep cortical layers of the rat barrel cortex. High spatial resolution EPI images could visualize intracortical macrovasculature and allowed the separation of voxels containing large intracortical vessels from surrounding voxels enriched with microvasculature. An early positive BOLD response was observed 0.8 s following stimulus onset mainly in tissue voxels. Around 1.2 s after stimulation, voxels
Conclusion
In this study, the evoked BOLD response can be detected as early as 0.8 s after the stimulus onset in the rodent barrel cortex. In contrast, the later onset of BOLD responses in individual macro vessels as compared to tissue voxels has been demonstrated. In addition, the peak BOLD signal of macro vessel voxels is at least twice as large as that of the tissue voxels. Thus, in addition to draining veins on the cortical surface, intracortical macro vessels can lead to mislocation of BOLD fMRI.
Acknowledgments
This research was supported by the Intramural Research Program of the NIH, NINDS. We thank Ms. Nadia Bouraoud and Ms. Kathryn Sharer for their technical support.
References (60)
- et al.
The neural basis of functional brain imaging signals
Trends Neurosci.
(2002) - et al.
In vivo micro-MRI of intracortical neurovasculature
NeuroImage
(2006) AFNI: software for analysis and visualization of functional magnetic resonance neuroimages
Comput. Biomed. Res.
(1996)- et al.
Temporal dynamics of the BOLD fMRI impulse response
NeuroImage
(2005) - et al.
Laminar specificity in monkey V1 using high-resolution SE-fMRI
Magn. Reson. Imaging
(2006) - et al.
Depth-resolved optical imaging and microscopy of vascular compartment dynamics during somatosensory stimulation
NeuroImage
(2007) - et al.
Spatial flow-volume dissociation of the cerebral microcirculatory response to mild hypercapnia
NeuroImage
(2006) - et al.
Cortical layer-dependent dynamic blood oxygenation, cerebral blood flow and cerebral blood volume responses during visual stimulation
NeuroImage
(2008) - et al.
High-resolution functional magnetic resonance imaging of the animal brain
Methods
(2003) - et al.
Spatial relationship between neuronal activity and BOLD functional MRI
NeuroImage
(2004)
Spatio-temporal point-spread function of fMRI signal in human gray matter at 7 Tesla
NeuroImage
Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field
Biochim. Biophys. Acta
How much cortex can a vein drain? Downstream dilution of activation-related cerebral blood oxygenation changes
NeuroImage
How accurate is magnetic resonance imaging of brain function?
Trends Neurosci.
An integrative model for neuronal activity-induced signal changes for gradient and spin echo functional imaging
NeuroImage
Signal and noise characteristics of Hahn SE and GE BOLD fMRI at 7 T in humans
NeuroImage
Robust detection of ocular dominance columns in humans using Hahn Spin Echo BOLD functional MRI at 7 Tesla
NeuroImage
3D mapping of somatotopic reorganization with small animal functional MRI
NeuroImage
Spatial specificity of cerebral blood volume-weighted fMRI responses at columnar resolution
NeuroImage
Glial and neuronal control of brain blood flow
Nature
Time course EPI of human brain function during task activation
Magn. Reson. Med.
Introduction to Functional Magnetic Resonance Imaging: Principles and Techniques
Mapping the body representation in the SI cortex of anesthetized and awake rats
J. Comp. Neurol.
Localized dynamic changes in cortical blood flow with whisker stimulation corresponds to matched vascular and neuronal architecture of rat barrels
J. Cereb. Blood Flow Metab.
Localized cerebral blood flow response at submillimeter columnar resolution
Proc. Natl. Acad. Sci. U. S. A.
Microvascular BOLD contribution at 4 and 7 T in the human brain: gradient-echo and spin-echo fMRI with suppression of blood effects
Magn. Reson. Med.
Inflow versus deoxyhemoglobin effects in BOLD functional MRI using gradient echoes at 1.5 T
NMR Biomed.
Retinotopic organization in human visual cortex and the spatial precision of functional MRI
Cereb. Cortex
Focal physiological uncoupling of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects
Proc. Natl. Acad. Sci. U. S. A.
Brief visual stimulation allows mapping of ocular dominance in visual cortex using fMRI
Hum. Brain Mapp.
Cited by (75)
Multimodal methods to help interpret resting-state fMRI
2023, Advances in Resting-State Functional MRI: Methods, Interpretation, and ApplicationsImaging faster neural dynamics with fast fMRI: A need for updated models of the hemodynamic response
2021, Progress in NeurobiologyCitation Excerpt :Comparing the HDR speed across brain regions highlights a potential trend in the speed of the HDR. Although data comparing the speed of the HDR across the brain are sparse, there is some preliminary evidence for a slowest HDR seen in the cerebral white matter (Li et al., 2019), then progressively faster responses seen in the cerebellar gray matter (Boillat and van der Zwaag, 2019; Chen and Desmond, 2005; Marvel and Desmond, 2012), followed by superficial then parenchymal cerebral gray matter (Kriegeskorte et al., 2010; Lewis et al., 2018b; Tian et al., 2010; Yu et al., 2012), subcortical gray matter such as the thalamus (Lau et al., 2011; Lewis et al., 2018b; Yen et al., 2011), then deeper subcortical gray matter in the brainstem (Faull and Pattinson, 2017; Lewis et al., 2018b), and potentially the fastest responses seen within the central nervous system just outside of the brain in the cervical spinal cord (Marcus et al., 1977; Nix et al., 1976; Piché et al., 2017; Sasaki et al., 2002). Although again some of these differences in the HDR across regions is likely attributable to differences in neuronal timing, the hypothesized progression of BOLD response delays across brain regions is summarized in Fig. 6.
Gastric stimulation drives fast BOLD responses of neural origin
2019, NeuroImage
- 1
Present address: Department of Electrical and Computer Engineering, Auburn University, Auburn, AL, 36894, USA.