Microstructural imaging of the human brain with a 'super-scanner': 10 key advantages of ultra-strong gradients for diffusion MRI.

The key component of a microstructural diffusion MRI 'super-scanner' is a dedicated high-strength gradient system that enables stronger diffusion weightings per unit time compared to conventional gradient designs. This can, in turn, drastically shorten the time needed for diffusion encoding, increase the signal-to-noise ratio, and facilitate measurements at shorter diffusion times. This review, written from the perspective of the UK National Facility for In Vivo MR Imaging of Human Tissue Microstructure, an initiative to establish a shared 300 mT/m-gradient facility amongst the microstructural imaging community, describes ten advantages of ultra-strong gradients for microstructural imaging. Specifically, we will discuss how the increase of the accessible measurement space compared to a lower-gradient systems (in terms of Δ, b-value, and TE) can accelerate developments in the areas of 1) axon diameter distribution mapping; 2) microstructural parameter estimation; 3) mapping micro-vs macroscopic anisotropy features with gradient waveforms beyond a single pair of pulsed-gradients; 4) multi-contrast experiments, e.g. diffusion-relaxometry; 5) tractography and high-resolution imaging in vivo and 6) post mortem; 7) diffusion-weighted spectroscopy of metabolites other than water; 8) tumour characterisation; 9) functional diffusion MRI; and 10) quality enhancement of images acquired on lower-gradient systems. We finally discuss practical barriers in the use of ultra-strong gradients, and provide an outlook on the next generation of 'super-scanners'.

Global signal modulation of single-trial fMRI response variability: Effect on positive vs negative BOLD response relationship.

In functional magnetic resonance imaging (fMRI), the relationship between positive BOLD responses (PBRs) and negative BOLD responses (NBRs) to stimulation is potentially informative about the balance of excitatory and inhibitory brain responses in sensory cortex. In this study, we performed three separate experiments delivering visual, motor or somatosensory stimulation unilaterally, to one side of the sensory field, to induce PBR and NBR in opposite brain hemispheres. We then assessed the relationship between the evoked amplitudes of contralateral PBR and ipsilateral NBR at the level of both single-trial and average responses. We measure single-trial PBR and NBR peak amplitudes from individual time-courses, and show that they were positively correlated in all experiments. In contrast, in the average response across trials the absolute magnitudes of both PBR and NBR increased with increasing stimulus intensity, resulting in a negative correlation between mean response amplitudes. Subsequent analysis showed that the amplitude of single-trial PBR was positively correlated with the BOLD response across all grey-matter voxels and was not specifically related to the ipsilateral sensory cortical response. We demonstrate that the global component of this single-trial response modulation could be fully explained by voxel-wise vascular reactivity, the BOLD signal standard deviation measured in a separate resting-state scan (resting state fluctuation amplitude, RSFA). However, bilateral positive correlation between PBR and NBR regions remained. We further report that modulations in the global brain fMRI signal cannot fully account for this positive PBR-NBR coupling and conclude that the local sensory network response reflects a combination of superimposed vascular and neuronal signals. More detailed quantification of physiological and noise contributions to the BOLD signal is required to fully understand the trial-by-trial PBR and NBR relationship compared with that of average responses.

Investigating intrinsic connectivity networks using simultaneous BOLD and CBF measurements.

When the sensory cortex is stimulated and directly receiving afferent input, modulations can also be observed in the activity of other brain regions comprising spatially distributed, yet intrinsically connected networks, suggesting that these networks support brain function during task performance. Such networks can exhibit subtle or unpredictable task responses which can pass undetected by conventional general linear modelling (GLM). Additionally, the metabolic demand of these networks in response to stimulation remains incompletely understood. Here, we recorded concurrent BOLD and CBF measurements during median nerve stimulation (MNS) and compared GLM analysis with independent component analysis (ICA) for identifying the spatial, temporal and metabolic properties of responses in the primary sensorimotor cortex (S1/M1), and in the default mode (DMN) and fronto-parietal (FPN) networks. Excellent spatial and temporal agreement was observed between the positive BOLD and CBF responses to MNS detected by GLM and ICA in contralateral S1/M1. Values of the change in cerebral metabolic rate of oxygen consumption (Δ%CMRO2) and the Δ%CMRO2/Δ%CBF coupling ratio were highly comparable when using either GLM analysis or ICA to extract the contralateral S1/M1 responses, validating the use of ICA for estimating changes in CMRO2. ICA identified DMN and FPN network activity that was not detected by GLM analysis. Using ICA, spatially coincident increases/decreases in both BOLD and CBF signals to MNS were found in the FPN/DMN respectively. Calculation of CMRO2 changes in these networks during MNS showed that the Δ%CMRO2/Δ%CBF ratio is comparable between the FPN and S1/M1 but is larger in the DMN than in the FPN, assuming an equal value of the parameter M in the DMN, FPN and S1/M1. This work suggests that metabolism-flow coupling may differ between these two fundamental brain networks, which could originate from differences between task-positive and task-negative fMRI responses, but might also be due to intrinsic differences between the two networks.

Evidence that the negative BOLD response is neuronal in origin: a simultaneous EEG-BOLD-CBF study in humans.

Unambiguous interpretation of changes in the BOLD signal is challenging because of the complex neurovascular coupling that translates changes in neuronal activity into the subsequent haemodynamic response. In particular, the neurophysiological origin of the negative BOLD response (NBR) remains incompletely understood. Here, we simultaneously recorded BOLD, EEG and cerebral blood flow (CBF) responses to 10 s blocks of unilateral median nerve stimulation (MNS) in order to interrogate the NBR. Both negative BOLD and negative CBF responses to MNS were observed in the same region of the ipsilateral primary sensorimotor cortex (S1/M1) and calculations showed that MNS induced a decrease in the cerebral metabolic rate of oxygen consumption (CMRO2) in this NBR region. The ∆CMRO2/∆CBF coupling ratio (n) was found to be significantly larger in this ipsilateral S1/M1 region (n=0.91±0.04, M=10.45%) than in the contralateral S1/M1 (n=0.65±0.03, M=10.45%) region that exhibited a positive BOLD response (PBR) and positive CBF response, and a consequent increase in CMRO2 during MNS. The fMRI response amplitude in ipsilateral S1/M1 was negatively correlated with both the power of the 8-13 Hz EEG mu oscillation and somatosensory evoked potential amplitude. Blocks in which the largest magnitude of negative BOLD and CBF responses occurred therefore showed greatest mu power, an electrophysiological index of cortical inhibition, and largest somatosensory evoked potentials. Taken together, our results suggest that a neuronal mechanism underlies the NBR, but that the NBR may originate from a different neurovascular coupling mechanism to the PBR, suggesting that caution should be taken in assuming the NBR simply represents the neurophysiological inverse of the PBR.

Spatial location and strength of BOLD activation in high-spatial-resolution fMRI of the motor cortex: a comparison of spin echo and gradient echo fMRI at 7 T.

The increased blood oxygenation level-dependent contrast-to-noise ratio at ultrahigh field (7 T) has been exploited in a comparison of the spatial location and strength of activation in high-resolution (1.5 mm isotropic) gradient echo (GE) and spin echo (SE), echo planar imaging data acquired during the execution of a simple motor task in five subjects. SE data were acquired at six echo times from 30 to 55 ms. Excellent fat suppression was achieved in the SE echo planar images using slice-selective gradient reversal. Threshold-free cluster enhancement was used to define regions of interest (ROIs) containing voxels showing significant stimulus-locked signal changes from the GE and average SE data. These were used to compare the signal changes and spatial locations of activated regions in SE and GE data. T(2) and T(2)* values were measured, with means of 48.3 ± 1.1 ms and 36.5 ± 3.4 ms in the SE ROI. In addition, we identified a dark band in SE images of the motor cortex corresponding to a region in which T(2) and T(2)* were significantly lower than in the surrounding grey matter. The fractional SE signal change in the ROI was found to vary linearly as a function of TE, with a slope that was dependent on the particular ROI assessed: the mean ΔR(2) value was found to be 0.85 ± 0.11 s(-1) for the SE ROI and -0.37 ± 0.05 s(-1) for the GE ROI. The fractional signal change relative to the shortest TE revealed that the largest signal change occurred at a TE of 45 ms outside of the dark band. At this TE, the ratio of the fractional signal change in GE and SE data was found to be 0.48 ± 0.05. Phase maps produced from high-resolution GE images spanning the right motor cortex were used to identify veins. The GE ROI was found to contain 18% more voxels overlying the venous mask than the SE ROI.

Bespoke magnetic field design for a magnetically shielded cold atom interferometer.

Quantum sensors based on cold atoms are being developed which produce measurements of unprecedented accuracy. Due to shifts in atomic energy levels, quantum sensors often have stringent requirements on their internal magnetic field environment. Typically, background magnetic fields are attenuated using high permeability magnetic shielding, with the cancelling of residual and introduction of quantisation fields implemented with coils inside the shield. The high permeability shield, however, distorts all magnetic fields, including those generated inside the sensor. Here, we demonstrate a solution by designing multiple coils overlaid on a 3D-printed former to generate three uniform and three constant linear gradient magnetic fields inside the capped cylindrical magnetic shield of a cold atom interferometer. The fields are characterised in-situ and match their desired forms to high accuracy. For example, the uniform transverse field, Bx, deviates by less than 0.2% over more than 40% of the length of the shield. We also map the field directly using the cold atoms and investigate the potential of the coil system to reduce bias from the quadratic Zeeman effect. This coil design technology enables targeted field compensation over large spatial volumes and has the potential to reduce systematic shifts and noise in numerous cold atom systems.

Investigating the effect of blood susceptibility on phase contrast in the human brain.

Recent work has shown a dramatic contrast between GM and WM in gradient echo phase images at high field (7 T). Although this contrast is key to the exploitation of phase in imaging normal and pathological tissue, its origin remains contentious. Several sources for this contrast have been considered including iron content, myelin, deoxy-hemoglobin, or water-macromolecule interactions. Here we quantify the contribution of intravascular dHb to the GM/WM contrast in the human brain at 7 T by modulating the susceptibility of the blood using a paramagnetic contrast agent. By carrying out high resolution, dynamic, gradient echo imaging before, during and after the injection of the contrast agent, we were able to follow the change in GM/WM phase contrast and to monitor simultaneously the susceptibility of the blood. Using these data in conjunction with the known susceptibility of venous blood we estimate the upper bound for the relative contribution of dHb in the vasculature to the measured GM/WM phase contrast to be 0.48 Hz for GM close to the pial surface, and 0.27 Hz for deeper GM. These values are up to 20% of the GM/WM phase difference observed in the human brain at 7 T. Furthermore, we found that the fractional blood volume differences required to account for the observed GM/WM phase contrast are 1.3% and 0.7% for GM close to the pial surface and for deeper GM, respectively.

Mapping human somatosensory cortex in individual subjects with 7T functional MRI.

Functional magnetic resonance imaging (fMRI) is now routinely used to map the topographic organization of human visual cortex. Mapping the detailed topography of somatosensory cortex, however, has proven to be more difficult. Here we used the increased blood-oxygen-level-dependent contrast-to-noise ratio at ultra-high field (7 Tesla) to measure the topographic representation of the digits in human somatosensory cortex at 1 mm isotropic resolution in individual subjects. A "traveling wave" paradigm was used to locate regions of cortex responding to periodic tactile stimulation of each distal phalangeal digit. Tactile stimulation was applied sequentially to each digit of the left hand from thumb to little finger (and in the reverse order). In all subjects, we found an orderly map of the digits on the posterior bank of the central sulcus (postcentral gyrus). Additionally, we measured event-related responses to brief stimuli for comparison with the topographic mapping data and related the fMRI responses to anatomical images obtained with an inversion-recovery sequence. Our results have important implications for the study of human somatosensory cortex and underscore the practical utility of ultra-high field functional imaging with 1 mm isotropic resolution for neuroscience experiments. First, topographic mapping of somatosensory cortex can be achieved in 20 min, allowing time for further experiments in the same session. Second, the maps are of sufficiently high resolution to resolve the representations of all five digits and third, the measurements are robust and can be made in an individual subject. These combined advantages will allow somatotopic fMRI to be used to measure the representation of digits in patients undergoing rehabilitation or plastic changes after peripheral nerve damage as well as tracking changes in normal subjects undergoing perceptual learning.

Measurement of electric fields induced in a human subject due to natural movements in static magnetic fields or exposure to alternating magnetic field gradients.

A dual dipole electric field probe has been used to measure surface electric fields in vivo on a human subject over a frequency range of 0.1-800 Hz. The low-frequency electric fields were induced by natural body movements such as walking and turning in the fringe magnetic fields of a 3 T magnetic resonance whole-body scanner. The rate-of-change of magnetic field (dB/dt) was also recorded simultaneously by using three orthogonal search coils positioned near to the location of the electric field probe. Rates-of-change of magnetic field for natural body rotations were found to exceed 1 T s(-1) near the end of the magnet bore. Typical electric fields measured on the upper abdomen, head and across the tongue for 1 T s(-1) rate of change of magnetic field were 0.15+/-0.02, 0.077+/-0.003 and 0.015+/-0.002 V m(-1) respectively. Electric fields on the abdomen and chest were measured during an echo-planar sequence with the subject positioned within the scanner. With the scanner rate-of-change of gradient set to 10 T m(-1) s(-1) the measured rate-of-change of magnetic field was 2.2+/-0.1 T s(-1) and the peak electric field was 0.30+/-0.01 V m(-1) on the chest. The values of induced electric field can be related to dB/dt by a 'geometry factor' for a given subject and sensor position. Typical values of this factor for the abdomen or chest (for measured surface electric fields) lie in the range of 0.10-0.18 m. The measured values of electric field are consistent with currently available numerical modelling results for movement in static magnetic fields and exposure to switched magnetic field gradients.

Measurement of electric fields due to time-varying magnetic field gradients using dipole probes.

The operation of dipole probes in measuring electric fields in conductive media exposed to temporally varying magnetic fields is discussed. The potential measured by the probe can be thought of as originating from two contributions to the electric field, namely the gradient of the scalar electric potential and the temporal derivative of the magnetic vector potential. Using this analysis, it is shown that the exact form of the wire paths employed when using electric field probes to measure the effects of temporally varying magnetic fields is very important and this prediction is verified via simple experiments carried out using different probe geometries in a cylindrical sample exposed to a temporally varying, uniform magnetic field. Extending this work, a dipole probe has been used to measure the electric field induced in a cylindrical sample by gradient coils as used in magnetic resonance imaging (MRI). Analytic solutions for the electric field in an infinite cylinder are verified by comparison with experimental measurements. Deviations from the analytic solutions of the electric field for the x-gradient coil due to the finite length of the sample cylinder are also demonstrated.

Theoretical optimization of multi-echo fMRI data acquisition.

Simulations are used to optimize multi-echo fMRI data acquisition for detection of BOLD signal changes in this study. Optimal sequence design (echo times and sampling period (receiver bandwidth)) and the variation in sensitivity between tissues with different baseline T*(2) are investigated, taking into account the effects of physiological noise and non-exponential signal decay. In the case of a single echo, for normally distributed, uncorrelated noise, the results indicate that the sampling period should be made as long as possible (so as to produce an acceptable level of image distortion), up to a maximum sampling period of 3T*(2), (i.e. optimum TE = 1.5T*(2)). Combining the signal from multiple echoes using weighted summation improves the contrast-to-noise ratio (CNR), at a reduced optimum echo interval. If the BOLD effect causes a constant change in relaxation rate, DeltaR*(2), independent of the tissue R*(2), then a multi-echo acquisition causes considerable variation in sensitivity to BOLD signal changes with tissue T*(2), so that if the sequence is optimized for a target tissue T*(2) it will be more sensitive to BOLD signal changes in tissues with shorter T*(2) values. Fitting for DeltaR*(2) reduces the CNR, and when using this approach, the shortest echo time interval should be used, down to a limit of about 0.3T*(2), and as many echoes as possible within the constraints of TR or hardware limitations should be collected. It is also shown that the optimal sequence will remain optimum or close to optimum irrespective of whether there are physiological noise contributions.

Sensory-specific satiety-related olfactory activation of the human orbitofrontal cortex.

When a food is eaten to satiety, its reward value decreases. This decrease is usually greater for the food eaten to satiety than for other foods, an effect termed sensory-specific satiety. In an fMRI investigation it was shown that for a region of the orbitofrontal cortex the activation produced by the odour of the food eaten to satiety decreased, whereas there was no similar decrease for the odour of a food not eaten in the meal. This effect was shown both by a voxel-wise SPM contrast (p <0.05 corrected) and an ANOVA performed on the mean percentage change in BOLD signal in the identified clusters of voxels (p <0.006). These results show that activation of a region of the human orbitofrontal cortex is related to olfactory sensory-specific satiety.

Sensory-specific satiety-related olfactory activation of the human orbitofrontal cortex.

When a food is eaten to satiety, its reward value decreases. This decrease is usually greater for the food eaten to satiety than for other foods, an effect termed sensory-specific satiety. In an fMRI investigation it was shown that for a region of the orbitofrontal cortex the activation produced by the odour of the food eaten to satiety decreased, whereas there was no similar decrease for the odour of a food not eaten in the meal. This effect was shown both by a voxel-wise SPM contrast (p<0.05 corrected) and an ANOVA performed on the mean percentage change in BOLD signal in the identified clusters of voxels (p<0.006). These results show that activation of a region of the human orbitofrontal cortex is related to olfactory sensory-specific satiety.

The representation of pleasant touch in the brain and its relationship with taste and olfactory areas.

Although there has been much investigation of brain pathways involved in pain, little is known about the brain mechanisms involved in processing somatosensory stimuli which feel pleasant. Employing fMRI it was shown that pleasant touch to the hand with velvet produced stronger activation of the orbitofrontal cortex than affectively neutral touch of the hand with wood. In contrast, the affectively neutral but more intense touch produced more activation of the primary somatosensory cortex than the pleasant stimulus. This indicates that part of the orbitofrontal cortex is concerned with representing the positively affective aspects of somatosensory stimuli, and in further experiments it was shown that this orbitofrontal area is different from that activated by taste and smell. The finding that three different primary or unlearned types of reinforcer (touch, taste, and smell) are represented in the orbitofrontal cortex helps to provide a firm foundation for understanding the neural basis of emotions, which can be understood in terms of states elicited by stimuli which are rewarding or punishing.

Resolution in high field echo planar microscopy.

The application of echo planar imaging to NMR microscopy offers a temporal resolution unparalleled by other techniques. However, a major difficulty in imaging at the high field strengths used for microscopy is the effect of local field inhomogeneities caused by magnetic susceptibility effects. This can give rise to both image distortion and signal loss. In addition, the effect of diffusion in the presence of the large imaging gradients gives rise to a broadening of the point spread function and hence loss of true resolution. We compare the sensitivity of two techniques, MBEST and PEPI, to both of these effects. Analytic expressions for the signal in each echo of the two sequences are developed, and the point spread functions for the two techniques are calculated. Using PEPI, we have been able to produce images with an in-plane resolution of 50 micrometer from a single free induction decay. This technique has been extended to three dimensions allowing the generation of 64(3) images with an isotropic resolution of 80 micrometer.

Imaging the long-range dipolar field in structured liquid state samples.

We describe imaging experiments in which the pattern of the dipolar field generated by spatially modulated nuclear magnetization is directly visualized in simply structured phantoms. Two types of experiment have been carried out at 11.7 T using (1)H NMR signals. In the first, the field from a single spin species is imaged via its own NMR signal. In the second, the NMR signal from one spin species is used to image the field generated by a second species. The field patterns measured in these experiments correspond well with those calculated using simple theoretical expressions for the dipolar field. The results also directly demonstrate the spatial sensitivity of the signal generated using dipolar field effects, indicating that the range of the field depends upon the inverse of the spatial frequency with which the magnetization is modulated.

Actively shielded multi-layer gradient coil designs with improved cooling properties.

In standard cylindrical gradient coils consisting of a single layer of wires, a limiting factor in achieving very large magnetic field gradients is the rapid increase in coil resistance with efficiency. This is a particular problem in small-bore scanners, such as those used for MR microscopy. By adopting a multi-layer design in which the coil wires are allowed to spread out into multiple layers wound at increasing radii, a more favourable scaling of resistance with efficiency is achieved, thus allowing the design of more powerful gradient coils with acceptable resistance values. Previously this approach has been applied to the design of unshielded, longitudinal, and transverse gradient coils. Here, the multi-layer approach has been extended to allow the design of actively shielded multi-layer gradient coils, and also to produce coils exhibiting enhanced cooling characteristics. An iterative approach to modelling the steady-state temperature distribution within the coil has also been developed. Results indicate that a good level of screening can be achieved in multi-layer coils, that small versions of such coils can yield higher efficiencies at fixed resistance than conventional two-layer (primary and screen) coils, and that performance improves as the number of layers of increases. Simulations show that by optimising multi-layer coils for cooling it is possible to achieve significantly higher gradient strengths at a fixed maximum operating temperature. A four-layer coil of 8 mm inner diameter has been constructed and used to test the steady-state temperature model.

Electric fields induced in a spherical volume conductor by temporally varying magnetic field gradients.

A homogeneous spherical volume conductor is used as a model system for the purpose of calculating electric fields induced in the human head by externally applied time-varying magnetic fields. We present results for the case where magnetic field gradient coils, used in magnetic resonance imaging (MRI), form the magnetic field, and we use these data to put limits on the rates of gradient change with time needed to produce nerve stimulation. The electric field is calculated analytically for the case of ideal longitudinal and transverse linear field gradients. We also show results from computer calculations yielding the electric field maps in a sphere when the field gradients are generated by a real MRI gradient coil set. In addition, the effect of shifting the sphere within each gradient coil volume is investigated. Numerical analysis shows similar results when applied to a model human head.

Strong gradients for spatially resolved diffusion measurements.

A new multilayer approach to gradient coil design, which allows the production of very strong gradient coils with reasonable resistance and consequent power dissipation, has been developed. Using this approach we have designed and built a strong z-gradient coil that will accommodate vertically mounted samples contained in 5-mm nuclear magnetic resonance tubes. The coil has an efficiency of 1.73 Tm-1A-1, an inductance of 49 microH, and a resistance of 1.8 omega, with a homogeneous volume consisting of a central cylinder of 4.5-mm length and diameter. This coil has been used to monitor the diffusion of water in Nylon 6.6 at room temperature, during desorption. This system is difficult to monitor via nuclear magnetic resonance (NMR), because the diffusion coefficients are typically less than 10(-13) m2s-1, while the T2 relaxation time is less than 1 ms even when the sample is fully saturated. The resulting measurements show a strong concentration dependence of the T2 relaxation time and self-diffusion coefficient of the absorbed water. The measured concentration profiles are consistent with a Fickian diffusion process with a concentration-dependent diffusion coefficient. The measured self-diffusion values are in reasonable agreement with those inferred from the variation of the concentration profiles as a function of time, using the one-dimensional Fickian diffusion equation.

Echo-planar microscopy of porous rocks.

A modified version of the echo-planar imaging sequence incorporating multiple 180 degrees RF pulses (PEPI) has been implemented at 11.7 T and used to generate high resolution images of porous solids. By preceding the PEPI sequence with an inversion recovery or flow-encoding sequence, rapid T1, and velocity mapping is possible in the microscopic domain.