The impact of vessel size, orientation and intravascular contribution on the neurovascular fingerprint of BOLD bSSFP fMRI

Submillimeter balanced SSFP BOLD-functional MRI accelerated with 3D stack-of-spirals at 9.4 T

PURPOSE

This work aims to improve the speed of balanced SSFP (bSSFP) acquisition with segmented 3D stack-of-spirals for functional brain studies at ultrahigh field.

METHODS

Functional experiments were performed with an accelerated 3D stack-of-spirals sequence with water excitation for fat suppression. The resulting data were reconstructed using an iterative algorithm with corrections for system imperfections such as trajectory deviations and B0 inhomogeneity. In the first set of experiments, we evaluated the signal change and stability with respect to echo and TR for a full-field checkerboard stimulus. To demonstrate the high spatio-temporal resolution of the developed method, the results of three optimized protocols at submillimeter resolution (0.6-mm isotropic and 0.8-mm isotropic) and at 1.2 mm isotropic resolution for whole-brain coverage were shown.

RESULTS

Water excitation and the model-based iterative reconstruction improved image quality. The BOLD-related signal changes increased with longer TE and longer TR. We observed an increase in thermal noise performance at lower TE and higher TR. However, signal stability deteriorates at higher TE and TR. Therefore, optimized protocols used shorter TE and moderately long TR to maximize the sensitivity and speed. Reproducible activations were detected along the gray-matter gyri in the submillimeter protocols with a median signal change of approximately 4% across subjects.

CONCLUSIONS

Three-dimensional stack-of-spirals enables passband balanced SSFP functional imaging at a much higher spatial and temporal scale, compared with conventional spoiled gradient-echo train sequences.

On the influence of the vascular architecture on Gradient Echo and Spin Echo BOLD fMRI signals across cortical depth: a simulation approach based on realistic 3D vascular networks

GE-BOLD contrast stands out as the predominant technique in functional MRI experiments for its high sensitivity and straightforward implementation. GE-BOLD exhibits rather similar sensitivity to vessels independent of their size at submillimeter resolution studies like those examining cortical columns and laminae. However, the presence of nonspecific macrovascular contributions poses a challenge to accurately isolate neuronal activity. SE-BOLD increases specificity towards small vessels, thereby enhancing its specificity to neuronal activity, due to the effective suppression of extravascular contributions caused by macrovessels with its refocusing pulse. However, even SE-BOLD measurements may not completely remove these macrovascular contributions. By simulating hemodynamic signals across cortical depth, we gain insights into vascular contributions to the laminar BOLD signal. In this study, we employed four realistic 3D vascular models to simulate oxygen saturation states in various vascular compartments, aiming to characterize both intravascular and extravascular contributions to GE and SE signals, and corresponding BOLD signal changes, across cortical depth at 7T. Simulations suggest that SE-BOLD cannot completely reduce the macrovascular contribution near the pial surface. Simulations also show that both the specificity and signal amplitude of BOLD signals at 7T depend on the spatial arrangement of large vessels throughout cortical depth and on the pial surface.

A fully synthetic three-dimensional human cerebrovascular model based on histological characteristics to investigate the hemodynamic fingerprint of the layer BOLD fMRI signal formation

Recent advances in functional magnetic resonance imaging (fMRI) at ultra-high field (≥7 tesla), novel hardware, and data analysis methods have enabled detailed research on neurovascular function, such as cortical layer-specific activity, in both human and nonhuman species. A widely used fMRI technique relies on the blood oxygen level-dependent (BOLD) signal. BOLD fMRI offers insights into brain function by measuring local changes in cerebral blood volume, cerebral blood flow, and oxygen metabolism induced by increased neuronal activity. Despite its potential, interpreting BOLD fMRI data is challenging as it is only an indirect measurement of neuronal activity. Computational modeling can help interpret BOLD data by simulating the BOLD signal formation. Current developments have focused on realistic 3D vascular models based on rodent data to understand the spatial and temporal BOLD characteristics. While such rodent-based vascular models highlight the impact of the angioarchitecture on the BOLD signal amplitude, anatomical differences between the rodent and human vasculature necessitate the development of human-specific models. Therefore, a computational framework integrating human cortical vasculature, hemodynamic changes, and biophysical properties is essential. Here, we present a novel computational approach: a three-dimensional VAscular MOdel based on Statistics (3D VAMOS), enabling the investigation of the hemodynamic fingerprint of the BOLD signal within a model encompassing a fully synthetic human 3D cortical vasculature and hemodynamics. Our algorithm generates microvascular and macrovascular architectures based on morphological and topological features from the literature on human cortical vasculature. By simulating specific oxygen saturation states and biophysical interactions, our framework characterizes the intravascular and extravascular signal contributions across cortical depth and voxel-wise levels for gradient-echo and spin-echo readouts. Thereby, the 3D VAMOS computational framework demonstrates that using human characteristics significantly affects the BOLD fingerprint, making it an essential step in understanding the fundamental underpinnings of layer-specific fMRI experiments.

Effect of in-plane and out-of-plane bifurcated microfluidic channels on the flow of aggregating red blood cells

The blood flow through our microvascular system is a renowned difficult process to understand because the complex flow behavior of blood is intertwined with the complex geometry it has to flow through. Conventional 2D microfluidics has provided important insights, but progress is hampered by the limitation of 2-D confinement. Here we use selective laser-induced etching to excavate non-planar 3-D microfluidic channels in glass that consist of two generations of bifurcations, heading towards more physiological geometries. We identify a cross-talk between the first and second bifurcation only when both bifurcations are in the same plane, as observed in 2D microfluidics. Contrarily, the flow in the branch where the second bifurcation is perpendicular to the first is hardly affected by the initial distortion. This difference in flow behavior is only observed when red blood cells are aggregated, due to the presence of dextran, and disappears by increasing the distance between both generations of bifurcations. Thus, 3-D structures scramble in-plane flow distortions, exemplifying the importance of experimenting with truly 3D microfluidic designs in order to understand complex physiological flow behavior.

Submillimeter fMRI Acquisition Techniques for Detection of Laminar and Columnar Level Brain Activation

Since the first demonstration in the early 1990s, functional MRI (fMRI) has emerged as one of the most powerful, noninvasive neuroimaging tools to probe brain functions. Subsequently, fMRI techniques have advanced remarkably, enabling the acquisition of functional signals with a submillimeter voxel size. This innovation has opened the possibility of investigating subcortical neural activities with respect to the cortical depths or cortical columns. For this purpose, numerous previous works have endeavored to design suitable functional contrast mechanisms and dedicated imaging techniques. Depending on the choice of the functional contrast, functional signals can be detected with high sensitivity or with improved spatial specificity to the actual activation site, and the pertaining issues have been discussed in a number of earlier works. This review paper primarily aims to provide an overview of the subcortical fMRI techniques that allow the acquisition of functional signals with a submillimeter resolution. Here, the advantages and disadvantages of the imaging techniques will be described and compared. We also summarize supplementary imaging techniques that assist in the analysis of the subcortical brain activation for more accurate mapping with reduced geometric deformation. This review suggests that there is no single universally accepted method as the gold standard for subcortical fMRI. Instead, the functional contrast and the corresponding readout imaging technique should be carefully determined depending on the purpose of the study. Due to the technical limitations of current fMRI techniques, most subcortical fMRI studies have only targeted partial brain regions. As a future prospect, the spatiotemporal resolution of fMRI will be pushed to satisfy the community's need for a deeper understanding of whole-brain functions and the underlying connectivity in order to achieve the ultimate goal of a time-resolved and layer-specific spatial scale. EVIDENCE LEVEL: 1 TECHNICAL EFFICACY: Stage 1.

A mechanistic computational framework to investigate the hemodynamic fingerprint of the blood oxygenation level-dependent signal

Blood oxygenation level-dependent (BOLD) functional magnetic resonance imaging (fMRI) is one of the most used imaging techniques to map brain activity or to obtain clinical information about human cortical vasculature, in both healthy and disease conditions. Nevertheless, BOLD fMRI is an indirect measurement of brain functioning triggered by neurovascular coupling. The origin of the BOLD signal is quite complex, and the signal formation thus depends, among other factors, on the topology of the cortical vasculature and the associated hemodynamic changes. To understand the hemodynamic evolution of the BOLD signal response in humans, it is beneficial to have a computational framework available that virtually resembles the human cortical vasculature, and simulates hemodynamic changes and corresponding MRI signal changes via interactions of intrinsic biophysical and magnetic properties of the tissues. To this end, we have developed a mechanistic computational framework that simulates the hemodynamic fingerprint of the BOLD signal based on a statistically defined, three-dimensional, vascular model that approaches the human cortical vascular architecture. The microvasculature is approximated through a Voronoi tessellation method and the macrovasculature is adapted from two-photon microscopy mice data. Using this computational framework, we simulated hemodynamic changes-cerebral blood flow, cerebral blood volume, and blood oxygen saturation-induced by virtual arterial dilation. Then we computed local magnetic field disturbances generated by the vascular topology and the corresponding blood oxygen saturation changes. This mechanistic computational framework also considers the intrinsic biophysical and magnetic properties of nearby tissue, such as water diffusion and relaxation properties, resulting in a dynamic BOLD signal response. The proposed mechanistic computational framework provides an integrated biophysical model that can offer better insights regarding the spatial and temporal properties of the BOLD signal changes.

Microvascular Specificity of Spin Echo BOLD fMRI: Impact of EPI Echo Train Length

A spatially specific fMRI acquisition requires specificity to the microvasculature that serves active neuronal sites. Macrovascular contributions will reduce the microvascular specificity but can be reduced by using spin echo (SE) sequences that use a π pulse to refocus static field inhomogeneities near large veins. The microvascular specificity of a SE-echo planar imaging (SE-EPI) scan depends on the echo train length (ETL)-duration, but the dependence is not well-characterized in humans at 7T. To determine how microvascular-specific SE-EPI BOLD is in humans at 7T, we developed a Monte Carlo voxel model that computes the signal of a proton ensemble residing in a vasculature subjected to a SE-EPI pulse sequence. We characterized the ETL-duration dependence of the microvascular specificity by simulating the BOLD signal as a function of ETL, the range adhering to experimentally realistic readouts. We performed a validation experiment for our simulation observations, in which we acquired a set of SE-EPI BOLD time series with varying ETL during a hyperoxic gas challenge. Both our simulations and measurements show an increase in macrovascular contamination as a function of ETL, with an increase of 30% according to our simulation and 60% according to our validation experiment between the shortest and longest ETL durations (23.1 - 49.7 ms). We conclude that the microvascular specificity decreases heavily with increasing ETL-durations. We recommend reducing the ETL-duration as much as possible to minimize macrovascular contamination in SE-EPI BOLD experiments. We additionally recommend scanning at high resolutions to minimize partial volume effects with CSF. CSF voxels show a large BOLD response, which can be attributed to both the presence of large veins (high blood volume) and molecular oxygen-induced T1-shortening (significant in a hyperoxia experiment). The magnified BOLD signal in a GM-CSF partial volume voxel reduces the desired microvascular specificity and, therefore, will hinder the interpretation of functional MRI activation patterns.

Cortical depth-dependent human fMRI of resting-state networks using EPIK

Introduction

Recent laminar-fMRI studies have substantially improved understanding of the evoked cortical responses in multiple sub-systems; in contrast, the laminar component of resting-state networks spread over the whole brain has been less studied due to technical limitations. Animal research strongly suggests that the supragranular layers of the cortex play a critical role in maintaining communication within the default mode network (DMN); however, whether this is true in this and other human cortical networks remains unclear.

Methods

Here, we used EPIK, which offers unprecedented coverage at sub-millimeter resolution, to investigate cortical broad resting-state dynamics with depth specificity in healthy volunteers.

Results

Our results suggest that human DMN connectivity is primarily supported by intermediate and superficial layers of the cortex, and furthermore, the preferred cortical depth used for communication can vary from one network to another. In addition, the laminar connectivity profile of some networks showed a tendency to change upon engagement in a motor task. In line with these connectivity changes, we observed that the amplitude of the low-frequency-fluctuations (ALFF), as well as the regional homogeneity (ReHo), exhibited a different laminar slope when subjects were either performing a task or were in a resting state (less variation among laminae, i.e., lower slope, during task performance compared to rest).

Discussion

The identification of varied laminar profiles concerning network connectivity, ALFF, and ReHo, observed across two brain states (task vs. rest) has major implications for the characterization of network-related diseases and suggests the potential diagnostic value of laminar fMRI in psychiatric disorders, e.g., to differentiate the cortical dynamics associated with disease stages linked, or not linked, to behavioral changes. The evaluation of laminar-fMRI across the brain encompasses computational challenges; nonetheless, it enables the investigation of a new dimension of the human neocortex, which may be key to understanding neurological disorders from a novel perspective.

Improved sensitivity and microvascular weighting of 3T laminar fMRI with GE-BOLD using NORDIC and phase regression

INTRODUCTION

Functional MRI with spatial resolution in the submillimeter domain enables measurements of activation across cortical layers in humans. This is valuable as different types of cortical computations, e.g., feedforward versus feedback related activity, take place in different cortical layers. Laminar fMRI studies have almost exclusively employed 7T scanners to overcome the reduced signal stability associated with small voxels. However, such systems are relatively rare and only a subset of those are clinically approved. In the present study, we examined if the feasibility of laminar fMRI at 3T could be improved by use of NORDIC denoising and phase regression.

METHODS

5 healthy subjects were scanned on a Siemens MAGNETOM Prisma 3T scanner. To assess across-session reliability, each subject was scanned in 3-8 sessions on 3-4 consecutive days. A 3D gradient echo EPI (GE-EPI) sequence was used for BOLD acquisitions (voxel size 0.82 mm isotopic, TR = 2.2 s) using a block design finger tapping paradigm. NORDIC denoising was applied to the magnitude and phase time series to overcome limitations in temporal signal-to-noise ratio (tSNR) and the denoised phase time series were subsequently used to correct for large vein contamination through phase regression.

RESULTS AND CONCLUSION

NORDIC denoising resulted in tSNR values comparable to or higher than commonly observed at 7T. Layer-dependent activation profiles could thus be extracted robustly, within and across sessions, from regions of interest located in the hand knob of the primary motor cortex (M1). Phase regression led to substantially reduced superficial bias in obtained layer profiles, although residual macrovascular contribution remained. We believe the present results support an improved feasibility of laminar fMRI at 3T.

The impact of vasomotion on analysis of rodent fMRI data

Introduction

Small animal fMRI is an essential part of translational research in the cognitive neurosciences. Due to small dimensions and animal physiology preclinical fMRI is prone to artifacts that may lead to misinterpretation of the data. To reach unbiased translational conclusions, it is, therefore, crucial to identify potential sources of experimental noise and to develop correction methods for contributions that cannot be avoided such as physiological noise. Aim of this study was to assess origin and prevalence of hemodynamic oscillations (HDO) in preclinical fMRI in rat, as well as their impact on data analysis.

Methods

Following the development of algorithms for HDO detection and suppression, HDO prevalence in fMRI measurements was investigated for different anesthetic regimens, comprising isoflurane and medetomidine, and for both gradient echo and spin echo fMRI sequences. In addition to assessing the effect of vasodilation on HDO, it was studied if HDO have a direct neuronal correlate using local field potential (LFP) recordings. Finally, the impact of HDO on analysis of fMRI data was assessed, studying both the impact on calculation of activation maps as well as the impact on brain network analysis. Overall, 303 fMRI measurements and 32 LFP recordings were performed in 71 rats.

Results

In total, 62% of the fMRI measurements showed HDO with a frequency of (0.20 ± 0.02) Hz. This frequent occurrence indicated that HDO cannot be generally neglected in fMRI experiments. Using the developed algorithms, HDO were detected with a specificity of 95%, and removed efficiently from the signal time courses. HDO occurred brain-wide under vasoconstrictive conditions in both small and large blood vessels. Vasodilation immediately interrupted HDO, which, however, returned within 1 h under vasoconstrictive conditions. No direct neuronal correlate of HDO was observed in LFP recordings. HDO significantly impacted analysis of fMRI data, leading to altered cluster sizes and F-values for activated voxels, as well as altered brain networks, when comparing data with and without HDO.

Discussion

We therefore conclude that HDO are caused by vasomotion under certain anesthetic conditions and should be corrected during fMRI data analysis to avoid bias.

Improved laminar specificity and sensitivity by combining SE and GE BOLD signals

The most widely used gradient-echo (GE) blood oxygenation level-dependent (BOLD) contrast has high sensitivity, but low specificity due to draining vein contributions, while spin-echo (SE) BOLD approach at ultra-high magnetic fields is highly specific to neural active sites but has lower sensitivity. To obtain high specificity and sensitivity, we propose to utilize a vessel-size-sensitive filter to the GE-BOLD signal, which suppresses macrovascular contributions and to combine selectively retained microvascular GE-BOLD signals with the SE-BOLD signals. To investigate our proposed idea, fMRI with 0.8 mm isotropic resolution was performed on the primary motor and sensory cortices in humans at 7 T by implementing spin- and gradient-echo (SAGE) echo planar imaging (EPI) acquisition. Microvascular-passed sigmoidal filters were designed based upon the vessel-size-sensitive ΔR₂*/ΔR₂ value for retaining GE-BOLD signals originating from venous vessels with ≤ 45 μm and ≤ 65 μm diameter. Unlike GE-BOLD fMRI, the laminar profile of SAGE-BOLD fMRI with the vessel-size-sensitive filter peaked at ∼ 1.0 mm from the surface of the primary motor and sensory cortices, demonstrating an improvement of laminar specificity over GE-BOLD fMRI. Also, the functional sensitivity of SAGE BOLD at middle layers (0.75-1.5 mm) was improved by ∼ 80% to ∼100% when compared with SE BOLD. In summary, we showed that combined GE- and SE-BOLD fMRI with the vessel-size-sensitive filter indeed yielded improved laminar specificity and sensitivity and is therefore an excellent tool for high spatial resolution ultra-high filed (UHF)-fMRI studies for resolving mesoscopic functional units.

Toward an integrative neurovascular framework for studying brain networks

Brain functional connectivity based on the measure of blood oxygen level-dependent (BOLD) functional magnetic resonance imaging (fMRI) signals has become one of the most widely used measurements in human neuroimaging. However, the nature of the functional networks revealed by BOLD fMRI can be ambiguous, as highlighted by a recent series of experiments that have suggested that typical resting-state networks can be replicated from purely vascular or physiologically driven BOLD signals. After going through a brief review of the key concepts of brain network analysis, we explore how the vascular and neuronal systems interact to give rise to the brain functional networks measured with BOLD fMRI. This leads us to emphasize a view of the vascular network not only as a confounding element in fMRI but also as a functionally relevant system that is entangled with the neuronal network. To study the vascular and neuronal underpinnings of BOLD functional connectivity, we consider a combination of methodological avenues based on multiscale and multimodal optical imaging in mice, used in combination with computational models that allow the integration of vascular information to explain functional connectivity.

Multi-echo balanced SSFP with a sequential phase-encoding order for functional MR imaging at 7T

PURPOSE

To develop a 2D multi-echo passband balanced SSFP (bSSFP) sequence using an echo-train readout with a sequential phase-encoding order (sequential multi-echo bSSFP), and evaluate its performance in fast functional brain imaging at 7 T.

METHODS

As images of sequential multi-echo bSSFP exhibit multiple ghosts due to periodic k-space modulations, a GRAPPA-based reconstruction method was proposed to eliminate ghosting artifacts. MRI experiments were performed to compare the image quality of multi-echo bSSFP and conventional single-echo bSSFP. Submillimeter-resolution fMRI using a checkerboard visual stimulus was conducted to compare the activation characteristics of multi-echo bSSFP, conventional single-echo bSSFP and standard gradient-echo EPI (GE-EPI).

RESULTS

A higher mean structural similarity index was found between images of single-echo bSSFP and multi-echo bSSFP with a shorter echo train length (ETL). Multi-echo bSSFP (ETL = 3) showed higher temporal SNR (tSNR) values than GRAPPA-accelerated single-echo bSSFP (R = 2). In submillimeter-resolution fMRI experiments, multi-echo bSSFP (ETL = 3) approached the imaging speed of GRAPPA-accelerated single-echo bSSFP (R = 2), but without tSNR penalty and reduced activation due to acceleration. The median t-value and the number of significantly activated voxels were comparable between GE-EPI and multi-echo bSSFP (ETL = 3) that provides virtually distortion-free functional images and inherits the activation patterns of conventional bSSFP.

CONCLUSION

Sequential multi-echo bSSFP (ETL = 3) is suitable for fast fMRI with submillimeter in-plane resolution, and offers an option to accelerate bSSFP imaging without tSNR penalty like parallel imaging.

Hemodynamic and metabolic changes during hypercapnia with normoxia and hyperoxia using pCASL and TRUST MRI in healthy adults

Blood oxygenation level-dependent (BOLD) or arterial spin labeling (ASL) MRI with hypercapnic stimuli allow for measuring cerebrovascular reactivity (CVR). Hypercapnic stimuli are also employed in calibrated BOLD functional MRI for quantifying neuronally-evoked changes in cerebral oxygen metabolism (CMRO₂). It is often assumed that hypercapnic stimuli (with or without hyperoxia) are iso-metabolic; increasing arterial CO₂ or O₂ does not affect CMRO₂. We evaluated the null hypothesis that two common hypercapnic stimuli, 'CO₂ in air' and carbogen, are iso-metabolic. TRUST and ASL MRI were used to measure the cerebral venous oxygenation and cerebral blood flow (CBF), from which the oxygen extraction fraction (OEF) and CMRO₂ were calculated for room-air, 'CO₂ in air' and carbogen. As expected, CBF significantly increased (9.9% ± 9.3% and 12.1% ± 8.8% for 'CO₂ in air' and carbogen, respectively). CMRO₂ decreased for 'CO₂ in air' (-13.4% ± 13.0%, p < 0.01) compared to room-air, while the CMRO₂ during carbogen did not significantly change. Our findings indicate that 'CO₂ in air' is not iso-metabolic, while carbogen appears to elicit a mixed effect; the CMRO₂ reduction during hypercapnia is mitigated when including hyperoxia. These findings can be important for interpreting measurements using hypercapnic or hypercapnic-hyperoxic (carbogen) stimuli.

Investigating mechanisms of fast BOLD responses: The effects of stimulus intensity and of spatial heterogeneity of hemodynamics

Recent studies have demonstrated that fast fMRI can track neural activity well above the temporal limit predicted by the canonical hemodynamic response model. While these findings are promising, the biophysical mechanisms underlying these fast fMRI phenomena remain underexplored. In this study, we discuss two aspects of the hemodynamic response, complementary to several existing hypotheses, that can accommodate faster fMRI dynamics beyond those predicted by the canonical model. First, we demonstrate, using both visual and somatosensory paradigms, that the timing and shape of hemodynamic response functions (HRFs) vary across graded levels of stimulus intensity-with lower-intensity stimulation eliciting faster and narrower HRFs. Second, we show that as the spatial resolution of fMRI increases, voxel-wise HRFs begin to deviate from the canonical model, with a considerable portion of voxels exhibiting faster temporal dynamics than predicted by the canonical HRF. Collectively, both stimulus/task intensity and image resolution can affect the sensitivity of fMRI to fast brain activity, which may partly explain recent observations of fast fMRI signals. It is further noteworthy that, while the present investigations focus on fast neural responses, our findings suggest that a revised hemodynamic model may benefit the many fMRI studies using paradigms with wide ranges of contrast levels (e.g., resting or naturalistic conditions) or with modern, high-resolution MR acquisitions.

Imaging faster neural dynamics with fast fMRI: A need for updated models of the hemodynamic response

Fast fMRI enables the detection of neural dynamics over timescales of hundreds of milliseconds, suggesting it may provide a new avenue for studying subsecond neural processes in the human brain. The magnitudes of these fast fMRI dynamics are far greater than predicted by canonical models of the hemodynamic response. Several studies have established nonlinear properties of the hemodynamic response that have significant implications for fast fMRI. We first review nonlinear properties of the hemodynamic response function that may underlie fast fMRI signals. We then illustrate the breakdown of canonical hemodynamic response models in the context of fast neural dynamics. We will then argue that the canonical hemodynamic response function is not likely to reflect the BOLD response to neuronal activity driven by sparse or naturalistic stimuli or perhaps to spontaneous neuronal fluctuations in the resting state. These properties suggest that fast fMRI is capable of tracking surprisingly fast neuronal dynamics, and we discuss the neuroscientific questions that could be addressed using this approach.

Improvement of sensitivity and specificity for laminar BOLD fMRI with double spin-echo EPI in humans at 7 T

Mapping mesoscopic cortical functional units such as columns or laminae is increasingly pursued by ultra-high field (UHF) functional magnetic resonance imaging (fMRI). The most popular approach for high-resolution fMRI is currently gradient-echo (GE) blood oxygenation level-dependent (BOLD) fMRI. However, its spatial accuracy is reduced due to its sensitivity to draining vessels, including pial veins, whereas spin-echo (SE) BOLD signal is expected to have higher spatial accuracy, albeit with lower sensitivity than the GE-BOLD signal. Here, we introduce a new double spin-echo (dSE) echo-planar imaging (EPI) method to improve the sensitivity of SE-BOLD contrast by averaging two spin-echoes using three radiofrequency pulses. Human fMRI experiments were performed with slices perpendicular to the central sulcus between motor and sensory cortices at 7 T during fist-clenching with touching. First, we evaluated the feasibility of single-shot dSE-EPI for BOLD fMRI with 1.5 mm isotropic resolution and found that dSE-BOLD fMRI has higher signal-to-noise ratio (SNR), temporal SNR (tSNR), and higher functional sensitivity than conventional SE-BOLD fMRI. Second, to investigate the laminar specificity of dSE-BOLD fMRI, we implemented a multi-shot approach to achieve 0.8-mm isotropic resolution with sliding-window reconstruction. Unlike GE-BOLD fMRI, the cortical profile of dSE-BOLD fMRI peaked at ~ 1.0 mm from the surface of the primary motor and sensory cortices, demonstrating an improvement of laminar specificity in humans over GE-BOLD fMRI. The proposed multi-shot dSE-EPI method is viable for high spatial resolution UHF-fMRI studies in the pursuit of resolving mesoscopic functional units.

Laminar fMRI using T2-prepared multi-echo FLASH

Functional magnetic resonance imaging (fMRI) using blood oxygenation level dependent (BOLD) contrast at a sub-millimeter scale is a promising technique to probe neural activity at the level of cortical layers. While gradient echo (GRE) BOLD sequences exhibit the highest sensitivity, their signal is confounded by unspecific extravascular (EV) and intravascular (IV) effects of large intracortical ascending veins and pial veins leading to a downstream blurring effect of local signal changes. In contrast, spin echo (SE) fMRI promises higher specificity towards signal changes near the microvascular compartment. However, the T₂-weighted signal is typically sampled with a gradient echo readout imposing additional T₂'-weighting. In this work, we used a T₂-prepared (T₂-prep) sequence with short GRE readouts to investigate its capability to acquire laminar fMRI data during a visual task in humans at 7 T. By varying the T₂-prep echo time (TE_prep) and acquiring multiple gradient echoes (TE_GRE) per excitation, we studied the specificity of the sequence and the influence of possible confounding contributions to the shape of laminar fMRI profiles. By fitting and extrapolating the multi-echo GRE data to a TE_GRE = 0 ms condition, we show for the first time laminar profiles free of T₂'-pollution, confined to gray matter. This finding is independent of TE_prep, except for the shortest one (31 ms) where hints of a remaining intravascular component can be seen. For TE_GRE > 0 ms a prominent peak at the pial surface is observed that increases with longer TE_GRE and dominates the shape of the profiles independent of the amount of T₂-weighting. Simulations show that the peak at the pial surface is a result of static EV dephasing around pial vessels in CSF visible in GM due to partial voluming. Additionally, another, weaker, static dephasing effect is observed throughout all layers of the cortex, which is particularly obvious in the data with shortest T₂-prep echo time. Our simulations show that this cannot be explained by intravascular dephasing but that it is likely caused by extravascular effects of the intracortical and pial veins. We conclude that even for TE_GRE as short as 2.3 ms, the T₂'-weighting added to the T₂-weighting is enough to dramatically affect the laminar specificity of the BOLD signal change. However, the bulk of this corruption stems from CSF partial volume effects which can in principle be addressed by increasing the spatial resolution of the acquisition.

Determining laminar neuronal activity from BOLD fMRI using a generative model

Laminar fMRI using the BOLD contrast enables the non-invasive investigation of mesoscopic functional circuits in the human brain. However, the laminar neuronal activity is spatiotemporally biased in the observed cortical depth profiles of the BOLD signal. In this study, we propose a generative fMRI signal model, comprehensively covering the relationship between cortical depth-dependent changes in excitatory and inhibitory neuronal activity with the sampling of the BOLD signal with finite voxels. The generative model allowed us to investigate pertinent questions regarding the accuracy of the laminar BOLD signal relative to the neuronal activity, and we found that: a) condition differences in laminar BOLD signals may be more reflective of neuronal activity than single condition BOLD signal depth profiles; b) angular dependence of the BOLD signal induces significant signal variability, which can mask underlying activity profiles; c) even if only three neuronal depths are of interest, more BOLD signal depths should be considered in the analysis. In addition, we recommend that the laminar BOLD data should be displayed using the centroid method to appreciate its spatial distribution in the original resolution. Finally, we showed that Bayesian model inversion of the generative model can improve sensitivity and specificity of assessing depth-dependent neuronal changes both for steady-state and dynamically.

Brain microvasculature has a common topology with local differences in geometry that match metabolic load

The microvasculature underlies the supply networks that support neuronal activity within heterogeneous brain regions. What are common versus heterogeneous aspects of the connectivity, density, and orientation of capillary networks? To address this, we imaged, reconstructed, and analyzed the microvasculature connectome in whole adult mice brains with sub-micrometer resolution. Graph analysis revealed common network topology across the brain that leads to a shared structural robustness against the rarefaction of vessels. Geometrical analysis, based on anatomically accurate reconstructions, uncovered a scaling law that links length density, i.e., the length of vessel per volume, with tissue-to-vessel distances. We then derive a formula that connects regional differences in metabolism to differences in length density and, further, predicts a common value of maximum tissue oxygen tension across the brain. Last, the orientation of capillaries is weakly anisotropic with the exception of a few strongly anisotropic regions; this variation can impact the interpretation of fMRI data.

Anatomical Modeling of Brain Vasculature in Two-Photon Microscopy by Generalizable Deep Learning

Objective and Impact Statement. Segmentation of blood vessels from two-photon microscopy (2PM) angiograms of brains has important applications in hemodynamic analysis and disease diagnosis. Here, we develop a generalizable deep learning technique for accurate 2PM vascular segmentation of sizable regions in mouse brains acquired from multiple 2PM setups. The technique is computationally efficient, thus ideal for large-scale neurovascular analysis. Introduction. Vascular segmentation from 2PM angiograms is an important first step in hemodynamic modeling of brain vasculature. Existing segmentation methods based on deep learning either lack the ability to generalize to data from different imaging systems or are computationally infeasible for large-scale angiograms. In this work, we overcome both these limitations by a method that is generalizable to various imaging systems and is able to segment large-scale angiograms. Methods. We employ a computationally efficient deep learning framework with a loss function that incorporates a balanced binary-cross-entropy loss and total variation regularization on the network’s output. Its effectiveness is demonstrated on experimentally acquired in vivo angiograms from mouse brains of dimensions up to 808×808×702 μm. Results. To demonstrate the superior generalizability of our framework, we train on data from only one 2PM microscope and demonstrate high-quality segmentation on data from a different microscope without any network tuning. Overall, our method demonstrates 10× faster computation in terms of voxels-segmented-per-second and 3× larger depth compared to the state-of-the-art. Conclusion. Our work provides a generalizable and computationally efficient anatomical modeling framework for brain vasculature, which consists of deep learning-based vascular segmentation followed by graphing. It paves the way for future modeling and analysis of hemodynamic response at much greater scales that were inaccessible before.

Effects of phase regression on high-resolution functional MRI of the primary visual cortex

High-resolution functional MRI studies have become a powerful tool to non-invasively probe the sub-millimeter functional organization of the human cortex. Advances in MR hardware, imaging techniques and sophisticated post-processing methods have allowed high resolution fMRI to be used in both the clinical and academic neurosciences. However, consensus within the community regarding the use of gradient echo (GE) or spin echo (SE) based acquisition remains largely divided. On one hand, GE provides a high temporal signal-to-noise ratio (tSNR) technique sensitive to both the macro- and micro-vascular signal while SE based methods are more specific to microvasculature but suffer from lower tSNR and specific absorption rate limitations, especially at high field and with short repetition times. Fortunately, the phase of the GE-EPI signal is sensitive to vessel size and this provides a potential avenue to reduce the macrovascular weighting of the signal (phase regression, Menon 2002). In order to determine the efficacy of this technique at high-resolution, phase regression was applied to GE-EPI timeseries and compared to SE-EPI to determine if GE-EPI's specificity to the microvascular compartment improved. To do this, functional data was collected from seven subjects on a neuro-optimized 7 T system at 800 μm isotropic resolution with both GE-EPI and SE-EPI while observing an 8 Hz contrast reversing checkerboard. Phase data from the GE-EPI was used to create a microvasculature-weighted time series (GE-EPI-PR). Anatomical imaging (MP2RAGE) was also collected to allow for surface segmentation so that the functional results could be projected onto a surface. A multi-echo gradient echo sequence was collected and used to identify venous vasculature. The GE-EPI-PR surface activation maps showed a high qualitative similarity with SE-EPI and also produced laminar activity profiles similar to SE-EPI. When the GE-EPI and GE-EPI-PR distributions were compared to SE-EPI it was shown that GE-EPI-PR had similar distribution characteristics to SE-EPI (p < 0.05) across the top 60% of cortex. Furthermore, it was shown that GE-EPI-PR has a higher contrast-to-noise ratio (0.5 ± 0.2, mean ± std. dev. across layers) than SE-EPI (0.27 ± 0.07) demonstrating the technique has higher sensitivity than SE-EPI. Taken together this evidence suggests phase regression is a useful method in low SNR studies such as high-resolution fMRI.

Anatomical Modeling of Brain Vasculature in Two-Photon Microscopy by Generalizable Deep Learning

Objective and Impact Statement. Segmentation of blood vessels from two-photon microscopy (2PM) angiograms of brains has important applications in hemodynamic analysis and disease diagnosis. Here, we develop a generalizable deep learning technique for accurate 2PM vascular segmentation of sizable regions in mouse brains acquired from multiple 2PM setups. The technique is computationally efficient, thus ideal for large-scale neurovascular analysis. Introduction. Vascular segmentation from 2PM angiograms is an important first step in hemodynamic modeling of brain vasculature. Existing segmentation methods based on deep learning either lack the ability to generalize to data from different imaging systems or are computationally infeasible for large-scale angiograms. In this work, we overcome both these limitations by a method that is generalizable to various imaging systems and is able to segment large-scale angiograms. Methods. We employ a computationally efficient deep learning framework with a loss function that incorporates a balanced binary-cross-entropy loss and total variation regularization on the network's output. Its effectiveness is demonstrated on experimentally acquired in vivo angiograms from mouse brains of dimensions up to 808 × 808 × 702   μ m . Results. To demonstrate the superior generalizability of our framework, we train on data from only one 2PM microscope and demonstrate high-quality segmentation on data from a different microscope without any network tuning. Overall, our method demonstrates 10× faster computation in terms of voxels-segmented-per-second and 3× larger depth compared to the state-of-the-art. Conclusion. Our work provides a generalizable and computationally efficient anatomical modeling framework for brain vasculature, which consists of deep learning-based vascular segmentation followed by graphing. It paves the way for future modeling and analysis of hemodynamic response at much greater scales that were inaccessible before.

Intravascular BOLD signal characterization of balanced SSFP experiments in human blood at high to ultrahigh fields

PURPOSE

To investigate the intravascular contribution to the overall balanced SSFP (bSSFP) BOLD effect in human blood at high to ultrahigh field strengths (3 T, 9.4 T, and 14.1 T).

METHODS

Venous blood prepared at two different oxygenation levels (deoxygenated: Y ≈ 71%, oxygenated: Y ≈ 94%) was measured with phase-cycled bSSFP for varying TRs/flip angles at 3 T, 9.4 T, and 14.1 T. The oxygen sensitivity was analyzed by intrinsic MIRACLE (motion-insensitive rapid configuration relaxometry)-R₂ estimation and passband signal differences. The intravascular BOLD-related signal change was extracted from the measured data for microvasculature and macrovasculature, and compared with the extravascular contribution obtained by Monte Carlo simulations.

RESULTS

The MIRACLE-R₂ values showed a characteristic increase with longer TRs in deoxygenated blood, corroborating that SE-R₂ data cannot be used to assess the intravascular bSSFP BOLD effect. Passband bSSFP signal differences measured at optimal flip angles of 30° at 3 T and 20° at 9.4 T/14.1 T revealed considerable relative intravascular contributions of 95%/70% at 3 T, 74%/43% at 9.4 T, 66%/46% at 14.1 T for TR = 5 ms, and 90%/65% at 3 T, 36%/27% at 9.4 T, 13%/15% at 14.1 T for TR = 10 ms in macrovascular/microvascular regimes.

CONCLUSION

The results indicate that intravascular effects have to be considered to better understand the origin of bSSFP BOLD contrast in functional MRI experiments, especially at short TRs. The MIRACLE-R₂ method demonstrated the ability to quantify the apparent decrease in R₂ due to rapid RF refocusing.

New acquisition techniques and their prospects for the achievable resolution of fMRI

This work reviews recent advances in technologies for functional magnetic resonance imaging (fMRI) of the human brain and highlights the push for higher functional specificity based on increased spatial resolution and specific MR contrasts to reveal previously undetectable functional properties of small-scale cortical structures. We discuss how the combination of MR hardware, advanced acquisition techniques and various MR contrast mechanisms have enabled recent progress in functional neuroimaging. However, these advanced fMRI practices have only been applied to a handful of neuroscience questions to date, with the majority of the neuroscience community still using conventional imaging techniques. We thus discuss upcoming challenges and possibilities for fMRI technology development in human neuroscience. We hope that readers interested in functional brain imaging acquire an understanding of current and novel developments and potential future applications, even if they don't have a background in MR physics or engineering. We summarize the capabilities of standard fMRI acquisition schemes with pointers to relevant literature and comprehensive reviews and introduce more recent developments.

High spatiotemporal vessel-specific hemodynamic mapping with multi-echo single-vessel fMRI

High-resolution fMRI enables noninvasive mapping of the hemodynamic responses from individual penetrating vessels in animal brains. Here, a 2D multi-echo single-vessel fMRI (MESV-fMRI) method has been developed to map the fMRI signal from arterioles and venules with a 100 ms sampling rate at multiple echo times (TE, 3-30 ms) and short acquisition windows (<1 ms). The T₂*-weighted signal shows the increased extravascular effect on venule voxels as a function of TE. In contrast, the arteriole voxels show an increased fMRI signal with earlier onset than venules voxels at the short TE (3 ms) with increased blood inflow and volume effects. MESV-fMRI enables vessel-specific T2* mapping and presents T2*-based fMRI time courses with higher contrast-to-noise ratios (CNRs) than the T2*-weighted fMRI signal at a given TE. The vessel-specific T2* mapping also allows semi-quantitative estimation of the oxygen saturation levels (Y) and their changes (ΔY) at a given blood volume fraction upon neuronal activation. The MESV-fMRI method enables vessel-specific T2* measurements with high spatiotemporal resolution for better modeling of the fMRI signal based on the hemodynamic parameters.

Modeling of vascular space occupancy and BOLD functional MRI from first principles using real microvascular angiograms

PURPOSE

The vascular space occupancy (VASO) is a functional MRI technique for probing cerebral blood volume changes noninvasively, including during neuronal activation in humans. An important consideration when implementing VASO is the BOLD effect in the signal. Assessing the physical origin of this BOLD contamination and the capabilities of correction methods could improve the quantification of cerebral blood volume changes with VASO.

METHODS

Given the heterogeneity of cerebral microvascular architecture, the vascular geometry within an MRI voxel can influence both BOLD and VASO signals. To investigate this effect, 3D high-resolution images of mouse cerebral vasculature measured with two-photon microscopy were used to model BOLD and VASO signals from first principles using Monte Carlo diffusion of water protons. Quantitative plots of VASO together with intravascular and extravascular BOLD signals as a function of TE at B₀ fields 1.5 T to 14 T were obtained.

RESULTS

The BOLD contamination of the VASO response was on the order of 50% for gradient echo and 5% for spin echo at 7 T and TE = 6 ms and significantly increased with TE and B₀ . Two currently used correction schemes were shown to account for most of this contamination and recover accurate relative signal changes, with optimal correction obtained using TEs as short as possible.

CONCLUSION

These results may provide useful information for optimizing sequence parameters in VASO and BOLD functional MRI, leading the way to a wider application of these techniques in healthy and diseased brain.

Comparison of BOLD and CBV using 3D EPI and 3D GRASE for cortical layer functional MRI at 7 T

PURPOSE

Functional MRI (fMRI) at the mesoscale of cortical layers and columns requires both sensitivity and specificity, the latter of which can be compromised if the imaging method is affected by vascular artifacts, particularly cortical draining veins at the pial surface. Recent studies have shown that cerebral blood volume (CBV) imaging is more specific to the actual laminar locus of neural activity than BOLD imaging using standard gradient-echo EPI sequences. Gradient and spin-echo (GRASE) BOLD imaging has also shown greater specificity when compared with standard gradient-echo EPI BOLD. Here we directly compare CBV and BOLD contrasts in high-resolution imaging of the primary motor cortex for laminar functional MRI in four combinations of signal labeling, CBV using slice-selective slab-inversion vascular space occupancy (VASO) and BOLD, each with 3D gradient-echo EPI and zoomed 3D-GRASE image readouts.

METHODS

Activations were measured using each sequence and contrast combination during a motor task. Activation profiles across cortical depth were measured to assess the sensitivity and specificity (pial bias) of each method.

RESULTS

Both CBV imaging using gradient-echo 3D-EPI and BOLD imaging using 3D-GRASE show similar specificity and sensitivity and are therefore useful tools for mesoscopic functional MRI in the human cortex. The combination of GRASE and VASO did not demonstrate high levels of sensitivity, nor show increased specificity.

CONCLUSION

Three-dimensional EPI with VASO contrast and 3D-GRASE with BOLD contrast both demonstrate sufficient sensitivity and specificity for laminar functional MRI to be used by neuroscientists in a wide range of investigations of depth-dependent neural circuitry in the human brain.

Robust functional mapping of layer-selective responses in human lateral geniculate nucleus with high-resolution 7T fMRI

The lateral geniculate nucleus (LGN) of the thalamus is the major subcortical relay of retinal input to the visual cortex. It plays important roles in visual perception and cognition and is closely related with several eye diseases and brain disorders. Primate LGNs mainly consist of six layers of monocular neurons with distinct cell types and functions. The non-invasive measure of layer-selective activities of the human LGN would have broad scientific and clinical implications. Using high-resolution functional magnetic resonance imaging (fMRI) at 7 Tesla (T) and carefully designed visual stimuli, we achieved robust functional mapping of eye-specific and also magnocellular/parvocellular-specific laminar patterns of the human LGN. These laminar patterns were highly reproducible with different pulse sequences scanned on separate days, between different subjects, and were in remarkable consistency with the simulation from high-resolution histology of the human LGNs. These findings clearly demonstrate that 7T fMRI can robustly resolve layer-specific responses of the human LGN. This paves the way for future investigation of the critical roles of the LGN in human visual perception and cognition, as well as the neural mechanisms of many developmental and neurodegenerative diseases.

Numerical simulation of the blood oxygenation level-dependent functional magnetic resonance signal using finite element method

Since the introduction of functional magnetic resonance imaging (fMRI), several computational approaches have been developed to examine the effect of the morphology and arrangement of blood vessels on the blood oxygenation-level dependent (BOLD) signal in the brain. In the present work, we implemented the original Ogawa's model using a numerical simulation based on the finite element method (FEM) instead of the analytical models. In literature, there are different works using analytical methods to analyse the transverse relaxation rate ( R2∗ ), which BOLD signal is related to, modelling the vascular system with simple and canonical geometries such as an infinite cylinder model (ICM) or a set of cylinders. We applied the numerical simulation to the extravascular BOLD signal as a function of angular vessel distribution (perpendicular vs parallel to the static magnetic field) relevant for anatomical districts characterized by geometrical symmetries, such as spinal cord. Numerical simulations confirmed analytical results for the canonical ICM. Moreover, the perturbation to the magnetic field induced by blood deoxyhaemoglobin, as quantified assuming Brownian diffusion of water molecules around the vessel, revealed that vessels contribute the most to the variation of the R2∗ when they are preferentially perpendicular to the external magnetic field, regardless of their size. Our results indicate that the numerical simulation method is sensitive to the effects of different vascular geometry. This work highlights the opportunity to extend R2∗ simulations to realistic models of vasculature based on high-resolution anatomical images.

Animal Functional Magnetic Resonance Imaging: Trends and Path Toward Standardization

Animal whole-brain functional magnetic resonance imaging (fMRI) provides a non-invasive window into brain activity. A collection of associated methods aims to replicate observations made in humans and to identify the mechanisms underlying the distributed neuronal activity in the healthy and disordered brain. Animal fMRI studies have developed rapidly over the past years, fueled by the development of resting-state fMRI connectivity and genetically encoded neuromodulatory tools. Yet, comparisons between sites remain hampered by lack of standardization. Recently, we highlighted that mouse resting-state functional connectivity converges across centers, although large discrepancies in sensitivity and specificity remained. Here, we explore past and present trends within the animal fMRI community and highlight critical aspects in study design, data acquisition, and post-processing operations, that may affect the results and influence the comparability between studies. We also suggest practices aimed to promote the adoption of standards within the community and improve between-lab reproducibility. The implementation of standardized animal neuroimaging protocols will facilitate animal population imaging efforts as well as meta-analysis and replication studies, the gold standards in evidence-based science.

A CMOS NMR needle for probing brain physiology with high spatial and temporal resolution

Magnetic resonance imaging and spectroscopy are versatile methods for probing brain physiology, but their intrinsically low sensitivity limits the achievable spatial and temporal resolution. Here, we introduce a monolithically integrated NMR-on-a-chip needle that combines an ultra-sensitive 300 µm NMR coil with a complete NMR transceiver, enabling in vivo measurements of blood oxygenation and flow in nanoliter volumes at a sampling rate of 200 Hz.

Automatic Graph-Based Modeling of Brain Microvessels Captured With Two-Photon Microscopy

Graph models of cerebral vasculature derived from two-photon microscopy have shown to be relevant to study brain microphysiology. Automatic graphing of these microvessels remain problematic due to the vascular network complexity and two-photon sensitivity limitations with depth. In this paper, we propose a fully automatic processing pipeline to address this issue. The modeling scheme consists of a fully-convolution neural network to segment microvessels, a three-dimensional surface model generator, and a geometry contraction algorithm to produce graphical models with a single connected component. Based on a quantitative assessment using NetMets metrics, at a tolerance of 60 μm, false negative and false positive geometric error 19 rates are 3.8% and 4.2%, respectively, whereas false nega- 20 tive and false positive topological error rates are 6.1% and 4.5%, respectively. Our qualitative evaluation confirms the efficiency of our scheme in generating useful and accurate graphical models.

Multi-slice passband bSSFP for human and rodent fMRI at ultra-high field

Balanced steady-state free precession (bSSFP) can be used as an alternative to gradient-echo (GE) EPI for BOLD functional MRI when image distortions and signal drop-outs are severe such as at ultra-high field. However, 3D-bSSFP acquisitions have distinct drawbacks on either human or animal MR systems. On clinical scanners, 3D imaging is suboptimal for localized fMRI applications. It can also display distortions when acceleration methods such as spiral read-outs are used, and, compared to multi-slice acquisitions, suffers from increased sensitivity to motion or physiological noise which further results in blurring. On pre-clinical systems, 3D acquisitions have low temporal resolution due to limited acceleration options, while single slice often results in insufficient coverage. The aim of the present study was to implement a multi-slice bSSFP acquisition with Cartesian read-out to obtain non-distorted BOLD fMRI activation maps in the human and rat brain at ultra-high field. We show that, when using a new pseudo-steady-state, the bSSFP signal characteristics are preserved. In the human brain at 7 T, we demonstrate that both task- and resting-state fMRI can be performed with multi-slice bSSFP, with a temporal SNR that matches that of 3D-bSSFP, resulting in - at least - equal performance. In the rat brain at 14 T, we show that the multi-slice bSSFP protocol has similar sensitivity to gradient-echo EPI for task fMRI, while benefitting from much reduced distortions and drop-outs. The advantages of passband bSSFP at 14 T in comparison with GE-EPI are expected to be even more marked for mouse brain.

Dependence of resting-state fMRI fluctuation amplitudes on cerebral cortical orientation relative to the direction of B0 and anatomical axes

Functional magnetic resonance imaging (fMRI) is now capable of sub-millimetre scale measurements over the entire human brain, however with such high resolutions each voxel is influenced by the local fine-scale details of the cerebral cortical vascular anatomy. The cortical vasculature is structured with the pial vessels lying tangentially along the grey matter surface, intracortical diving arterioles and ascending venules running perpendicularly to the surface, and a randomly oriented capillary network within the parenchyma. It is well-known that the amplitude of the blood-oxygenation level dependent (BOLD) signal emanating from a vessel depends on its orientation relative to the B₀-field. Thus the vascular geometric hierarchy will impart an orientation dependence to the BOLD signal amplitudes and amplitude differences due to orientation differences constitute a bias for interpreting neuronal activity. Here, we demonstrate a clear effect of cortical orientation to B₀ in the resting-state BOLD-fMRI amplitude (quantified as the coefficient of temporal signal variation) for 1.1 mm isotropic data at 7T and 2 mm isotropic at 3T. The maximum bias, i.e. the fluctuation amplitude difference between regions where cortex is perpendicular to vs. parallel to B₀, is about +70% at the pial surface at 7T and +11% at 3T. The B₀ orientation bias declines with cortical depth, becomes progressively smaller closer to the white matter surface, but then increases again to a local maximum within the white matter just beneath the cortical grey matter, suggesting a distinct tangential network of white matter vessels that also generate a BOLD orientation effect. We further found significant (negative) biases with the cortex orientation to the anterior-posterior anatomical axis of the head: a maximum negative bias of about -30% at the pial surface at 7T and about -13% at 3T. The amount of signal variance explained by the low frequency drift, motion and the respiratory cycle also showed a cortical orientation dependence; only the cardiac cycle induced signal variance was independent of cortical orientation, suggesting that the cardiac induced component of the image time-series fluctuations is not related to a significant change in susceptibility. Although these orientation effects represent a signal bias, and are likely to be a nuisance in high-resolution analyses, they may help characterize the vascular influences on candidate fMRI acquisitions and, thereby, may be exploited to improve the neuronal specificity of fMRI.

Dependence of the MR signal on the magnetic susceptibility of blood studied with models based on real microvascular networks

PURPOSE

The primary goal of this study was to estimate the value of β , the exponent in the power law relating changes of the transverse relaxation rate and intra-extravascular local magnetic susceptibility differences as Δ R 2 ∗ ∝ ( Δ χ ) β . The secondary objective was to evaluate any differences that might exist in the value of β obtained using a deoxyhemoglobin-weighted Δ χ distribution versus a constant Δ χ distribution assumed in earlier computations. The third objective was to estimate the value of β that is relevant for methods based on susceptibility contrast agents with a concentration of Δ χ higher than that used for BOLD fMRI calculations.

METHODS

Our recently developed model of real microvascular anatomical networks is used to extend the original simplified Monte-Carlo simulations to compute β from the first principles.

RESULTS

Our results show that β = 1 for most BOLD fMRI measurements of real vascular networks, as opposed to earlier predictions of β = 1 .5 using uniform Δ χ distributions. For perfusion or fMRI methods based on contrast agents, which generate larger values for Δ χ , β = 1 for B 0 ≤ 9.4 T, whereas at 14 T β can drop below 1 and the variation across subjects is large, indicating that a lower concentration of contrast agent with a lower value of Δ χ is desired for experiments at high B₀ .

CONCLUSION

These results improve our understanding of the relationship between R₂ ^* and the underlying microvascular properties. The findings will help to infer the cerebral metabolic rate of oxygen and cerebral blood volume from BOLD and perfusion MRI, respectively.