Thermal stimulus task fMRI in the cervical spinal cord at 7 Tesla.

Although functional magnetic resonance imaging (fMRI) is widely applied in the brain, fMRI of the spinal cord is more technically demanding. Proximity to the vertebral column and lungs results in strong spatial inhomogeneity and temporal fluctuations in B0 . Increasing field strength enables higher spatial resolution and improved sensitivity to blood oxygenation level-dependent (BOLD) signal, but amplifies the effects of B0 inhomogeneity. In this work, we present the first task fMRI in the spinal cord at 7 T. Further, we compare the performance of single-shot and multi-shot 2D echo-planar imaging (EPI) protocols, which differ in sensitivity to spatial and temporal B0 inhomogeneity. The cervical spinal cords of 11 healthy volunteers were scanned at 7 T using single-shot 2D EPI at 0.75 mm in-plane resolution and multi-shot 2D EPI at 0.75 and 0.6 mm in-plane resolutions. All protocols used 3 mm slice thickness. For each protocol, the BOLD response to 13 10-s noxious thermal stimuli applied to the right thumb was acquired in a 10-min fMRI run. Image quality, temporal signal to noise ratio (SNR), and BOLD activation (percent signal change and z-stat) at both individual- and group-level were evaluated between the protocols. Temporal SNR was highest in single-shot and multi-shot 0.75 mm protocols. In group-level analyses, activation clusters appeared in all protocols in the ipsilateral dorsal quadrant at the expected C6 neurological level. In individual-level analyses, activation clusters at the expected level were detected in some, but not all subjects and protocols. Single-shot 0.75 mm generally produced the highest mean z-statistic, while multi-shot 0.60 mm produced the best-localized activation clusters and the least geometric distortion. Larger than expected within-subject segmental variation of BOLD activation along the cord was observed. Group-level sensory task fMRI of the cervical spinal cord is feasible at 7 T with single-shot or multi-shot EPI. The best choice of protocol will likely depend on the relative importance of sensitivity to activation versus spatial localization of activation for a given experiment. PRACTITIONER POINTS: First stimulus task fMRI results in the spinal cord at 7 T. Single-shot 0.75 mm 2D EPI produced the highest mean z-statistic. Multi-shot 0.60 mm 2D EPI provided the best-localized activation and least distortion.

HiHi fMRI: a data-reordering method for measuring the hemodynamic response of the brain with high temporal resolution and high SNR.

There is emerging evidence that sampling the blood-oxygen-level-dependent (BOLD) response with high temporal resolution opens up new avenues to study the in vivo functioning of the human brain with functional magnetic resonance imaging. Because the speed of sampling and the signal level are intrinsically connected in magnetic resonance imaging via the T1 relaxation time, optimization efforts usually must make a trade-off to increase the temporal sampling rate at the cost of the signal level. We present a method, which combines a sparse event-related stimulus paradigm with subsequent data reshuffling to achieve high temporal resolution while maintaining high signal levels (HiHi). The proof-of-principle is presented by separately measuring the single-voxel time course of the BOLD response in both the primary visual and primary motor cortices with 100-ms temporal resolution.

Reliability of resting-state functional connectivity in the human spinal cord: Assessing the impact of distinct noise sources.

The investigation of spontaneous fluctuations of the blood-oxygen-level-dependent (BOLD) signal has recently been extended from the brain to the spinal cord, where it has stimulated interest from a clinical perspective. A number of resting-state functional magnetic resonance imaging (fMRI) studies have demonstrated robust functional connectivity between the time series of BOLD fluctuations in bilateral dorsal horns and between those in bilateral ventral horns, in line with the functional neuroanatomy of the spinal cord. A necessary step prior to extension to clinical studies is assessing the reliability of such resting-state signals, which we aimed to do here in a group of 45 healthy young adults at the clinically prevalent field strength of 3T. When investigating connectivity in the entire cervical spinal cord, we observed fair to good reliability for dorsal-dorsal and ventral-ventral connectivity, whereas reliability was poor for within- and between-hemicord dorsal-ventral connectivity. Considering how prone spinal cord fMRI is to noise, we extensively investigated the impact of distinct noise sources and made two crucial observations: removal of physiological noise led to a reduction in functional connectivity strength and reliability - due to the removal of stable and participant-specific noise patterns - whereas removal of thermal noise considerably increased the detectability of functional connectivity without a clear influence on reliability. Finally, we also assessed connectivity within spinal cord segments and observed that while the pattern of connectivity was similar to that of whole cervical cord, reliability at the level of single segments was consistently poor. Taken together, our results demonstrate the presence of reliable resting-state functional connectivity in the human spinal cord even after thoroughly accounting for physiological and thermal noise, but at the same time urge caution if focal changes in connectivity (e.g. due to segmental lesions) are to be studied, especially in a longitudinal manner.

Automated slice-specific z-shimming for functional magnetic resonance imaging of the human spinal cord.

Functional magnetic resonance imaging (fMRI) of the human spinal cord faces many challenges, such as signal loss due to local magnetic field inhomogeneities. This issue can be addressed with slice-specific z-shimming, which compensates for the dephasing effect of the inhomogeneities using a slice-specific gradient pulse. Here, we aim to address outstanding issues regarding this technique by evaluating its effects on several aspects that are directly relevant for spinal fMRI and by developing two automated procedures in order to improve upon the time-consuming and subjective nature of manual selection of z-shims: one procedure finds the z-shim that maximizes signal intensity in each slice of an EPI reference-scan and the other finds the through-slice field inhomogeneity for each EPI-slice in field map data and calculates the required compensation gradient moment. We demonstrate that the beneficial effects of z-shimming are apparent across different echo times, hold true for both the dorsal and ventral horn, and are also apparent in the temporal signal-to-noise ratio (tSNR) of EPI time-series data. Both of our automated approaches were faster than the manual approach, lead to significant improvements in gray matter tSNR compared to no z-shimming and resulted in beneficial effects that were stable across time. While the field-map-based approach performed slightly worse than the manual approach, the EPI-based approach performed as well as the manual one and was furthermore validated on an external corticospinal data-set (N > 100). Together, automated z-shimming may improve the data quality of future spinal fMRI studies and lead to increased reproducibility in longitudinal studies.

Feasibility of spiral fMRI based on an LTI gradient model.

Spiral imaging is very well suited for functional MRI, however its use has been limited by the fact that artifacts caused by gradient imperfections and B0 inhomogeneity are more difficult to correct compared to EPI. Effective correction requires accurate knowledge of the traversed k-space trajectory. With the goal of making spiral fMRI more accessible, we have evaluated image reconstruction using trajectories predicted by the gradient impulse response function (GIRF), which can be determined in a one-time calibration step. GIRF-predicted reconstruction was tested for high-resolution (0.8 mm) fMRI at 7T. Image quality and functional results of the reconstructions using GIRF-prediction were compared to reconstructions using the nominal trajectory and concurrent field monitoring. The reconstructions using nominal spiral trajectories contain substantial artifacts and the activation maps contain misplaced activation. Image artifacts are substantially reduced when using the GIRF-predicted reconstruction, and the activation maps for the GIRF-predicted and monitored reconstructions largely overlap. The GIRF reconstruction provides a large increase in the spatial specificity of the activation compared to the nominal reconstruction. The GIRF-reconstruction generates image quality and fMRI results similar to using a concurrently monitored trajectory. The presented approach does not prolong or complicate the fMRI acquisition. Using GIRF-predicted trajectories has the potential to enable high-quality spiral fMRI in situations where concurrent trajectory monitoring is not available.

A method for correcting breathing-induced field fluctuations in T2*-weighted spinal cord imaging using a respiratory trace.

PURPOSE: Spinal cord MRI at ultrahigh field is hampered by time-varying magnetic fields associated with the breathing cycle, giving rise to ghosting artifacts in multi-shot acquisitions. Here, we suggest a correction approach based on linking the signal from a respiratory bellows to field changes inside the spinal cord. The information is used to correct the data at the image reconstruction level. METHODS: The correction was demonstrated in the context of multi-shot T2*-weighted imaging of the cervical spinal cord at 7T. A respiratory trace was acquired during a high-resolution multi-echo gradient-echo sequence, used for structural imaging and quantitative T2* mapping, and a multi-shot EPI time series, as would be suitable for fMRI. The coupling between the trace and the breathing-induced fields was determined by a short calibration scan in each individual. Images were reconstructed with and without trace-based correction. RESULTS: In the multi-echo acquisition, breathing-induced fields caused severe ghosting in images with long TE, which led to a systematic underestimation of T2* in the spinal cord. The trace-based correction reduced the ghosting and increased the estimated T2* values. Breathing-related ghosting was also observed in the multi-shot EPI images. The correction largely removed the ghosting, thereby improving the temporal signal-to-noise ratio of the time series. CONCLUSIONS: Trace-based retrospective correction of breathing-induced field variations can reduce ghosting and improve quantitative metrics in multi-shot structural and functional T2*-weighted imaging of the spinal cord. The method is straightforward to implement and does not rely on sequence modifications or additional hardware beyond a respiratory bellows.

Spatiotemporal characterization of breathing-induced B0 field fluctuations in the cervical spinal cord at 7T.

Magnetic resonance imaging and spectroscopy of the spinal cord stand to benefit greatly from the increased signal-to-noise ratio of ultra-high field. However, ultra-high field also poses considerable technical challenges, especially related to static and dynamic B0 fields. Breathing causes the field to fluctuate with the respiratory cycle, giving rise to artifacts such as ghosting and apparent motion in images. We here investigated the spatial and temporal characteristics of breathing-induced B0 fields in the cervical spinal cord at 7T. We analyzed the magnitude and spatial profile of breathing-induced fields during breath-holds in an expired and inspired breathing state. We also measured the temporal field evolution during free breathing by acquiring a time series of fast phase images, and a principal component analysis was performed on the measured field evolution. In all subjects, the field shift was largest around the vertebral level of C7 and lowest at the top of the spinal cord. At C7, we measured peak-to-peak field fluctuations of 36 Hz on average during normal free breathing; increasing to on average 113 Hz during deep breathing. The first principal component could explain more than 90% of the field variations along the foot-head axis inside the spinal cord in all subjects. We further implemented a proof-of-principle shim correction, demonstrating the feasibility of using the shim system to compensate for the breathing-induced fields inside the spinal cord. Effective correction strategies will be crucial to unlock the full potential of ultra-high field for spinal cord imaging.

Spinal cord MRI at 7T.

Magnetic resonance imaging (MRI) of the human spinal cord at 7T has been demonstrated by a handful of research sites worldwide, and the spinal cord remains one of the areas in which higher fields and resolution could have high impact. The small diameter of the cord (∼1 cm) necessitates high spatial resolution to minimize partial volume effects between gray and white matter, and so MRI of the cord can greatly benefit from increased signal-to-noise ratio and contrasts at ultra-high field (UHF). Herein we review the current state of UHF spinal cord imaging. Technical challenges to successful UHF spinal cord MRI include radiofrequency (B1) nonuniformities and a general lack of optimized radiofrequency coils, amplified physiological noise, and an absence of methods for robust B0 shimming along the cord to mitigate image distortions and signal losses. Numerous solutions to address these challenges have been and are continuing to be explored, and include novel approaches for signal excitation and acquisition, dynamic shimming and specialized shim coils, and acquisitions with increased coverage or optimal slice angulations.

Template-based field map prediction for rapid whole brain B0 shimming.

PURPOSE: In typical MRI protocols, time is spent acquiring a field map to calculate the shim settings for best image quality. We propose a fast template-based field map prediction method that yields near-optimal shims without measuring the field. METHODS: The template-based prediction method uses prior knowledge of the B0 distribution in the human brain, based on a large database of field maps acquired from different subjects, together with subject-specific structural information from a quick localizer scan. The shimming performance of using the template-based prediction is evaluated in comparison to a range of potential fast shimming methods. RESULTS: Static B0 shimming based on predicted field maps performed almost as well as shimming based on individually measured field maps. In experimental evaluations at 7 T, the proposed approach yielded a residual field standard deviation in the brain of on average 59 Hz, compared with 50 Hz using measured field maps and 176 Hz using no subject-specific shim. CONCLUSIONS: This work demonstrates that shimming based on predicted field maps is feasible. The field map prediction accuracy could potentially be further improved by generating the template from a subset of subjects, based on parameters such as head rotation and body mass index. Magn Reson Med 80:171-180, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Analysis and correction of field fluctuations in fMRI data using field monitoring.

This work investigates the role of magnetic field fluctuations as a confound in fMRI. In standard fMRI experiments with single-shot EPI acquisition at 3 Tesla the uniform and gradient components of the magnetic field were recorded with NMR field sensors. By principal component analysis it is found that differences of field evolution between the EPI readouts are explainable by few components relating to slow and within-shot field dynamics of hardware and physiological origin. The impact of fluctuating field components is studied by selective data correction and assessment of its influence on image fluctuation and SFNR. Physiological field fluctuations, attributed to breathing, were found to be small relative to those of hardware origin. The dominant confounds were hardware-related and attributable to magnet drift and thermal changes. In raw image time series, field fluctuation caused significant SFNR loss, reflected by a 67% gain upon correction. Large part of this correction can be accomplished by traditional image realignment, which addresses slow and spatially uniform field changes. With realignment, explicit field correction increased the SFNR on the order of 6%. In conclusion, field fluctuations are a relevant confound in fMRI and can be addressed effectively by retrospective data correction. Based on the physics involved it is anticipated that the advantage of full field correction increases with field strength, with non-Cartesian readouts, and upon phase-sensitive BOLD analysis.

Gradient and shim pre-emphasis by inversion of a linear time-invariant system model.

PURPOSE: The goal of this contribution is to enhance the fidelity and switching speed of gradient and shim fields by advancing pre-emphasis toward broadband and full cross-term correction. THEORY AND METHODS: The proposed approach is based on viewing gradient and shim chains as linear, time-invariant (LTI) systems. Pre-emphasis is accomplished by inversion of a broadband digital system model. In the multiple-channel case, it amounts to a matrix of broadband filters that perform concerted self- and cross-term correction. This approach is demonstrated with gradients and shims up to the third order in a 7 Tesla whole-body MR system. RESULTS: Pre-emphasis by LTI model inversion is first verified by studying settling speeds and response behavior without and with the correction. It is then demonstrated for rapid shim updating, achieving substantially enhanced fidelity of field dynamics and shim settling within approximately 1 ms. In single-shot echo-planar imaging (EPI) acquisitions in vivo, this benefit is shown to translate into enhanced geometric fidelity. CONCLUSIONS: The fidelity of gradient and shim dynamics can be greatly enhanced by pre-emphasis based on inverting a general LTI system model. Permitting shim settling on the millisecond scale, broadband multiple-channel pre-emphasis promises to render higher-order shimming fully versatile at the level of MRI sequence design. Magn Reson Med 78:1607-1622, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Single-shot spiral imaging enabled by an expanded encoding model: Demonstration in diffusion MRI.

PURPOSE: The purpose of this work was to improve the quality of single-shot spiral MRI and demonstrate its application for diffusion-weighted imaging. METHODS: Image formation is based on an expanded encoding model that accounts for dynamic magnetic fields up to third order in space, nonuniform static B0 , and coil sensitivity encoding. The encoding model is determined by B0 mapping, sensitivity mapping, and concurrent field monitoring. Reconstruction is performed by iterative inversion of the expanded signal equations. Diffusion-tensor imaging with single-shot spiral readouts is performed in a phantom and in vivo, using a clinical 3T instrument. Image quality is assessed in terms of artefact levels, image congruence, and the influence of the different encoding factors. RESULTS: Using the full encoding model, diffusion-weighted single-shot spiral imaging of high quality is accomplished both in vitro and in vivo. Accounting for actual field dynamics, including higher orders, is found to be critical to suppress blurring, aliasing, and distortion. Enhanced image congruence permitted data fusion and diffusion tensor analysis without coregistration. CONCLUSION: Use of an expanded signal model largely overcomes the traditional vulnerability of spiral imaging with long readouts. It renders single-shot spirals competitive with echo-planar readouts and thus deploys shorter echo times and superior readout efficiency for diffusion imaging and further prospective applications. Magn Reson Med 77:83-91, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Image reconstruction using a gradient impulse response model for trajectory prediction.

PURPOSE: Gradient imperfections remain a challenge in MRI, especially for sequences relying on long imaging readouts. This work aims to explore image reconstruction based on k-space trajectories predicted by an impulse response model of the gradient system. THEORY AND METHODS: Gradient characterization was performed twice with 3 years interval on a commercial 3 Tesla (T) system. The measured gradient impulse response functions were used to predict actual k-space trajectories for single-shot echo-planar imaging (EPI), spiral and variable-speed EPI sequences. Image reconstruction based on the predicted trajectories was performed for phantom and in vivo data. Resulting images were compared with reconstructions based on concurrent field monitoring, separate trajectory measurements, and nominal trajectories. RESULTS: Image reconstruction using model-based trajectories yielded high-quality images, comparable to using separate trajectory measurements. Compared with using nominal trajectories, it strongly reduced ghosting, blurring, and geometric distortion. Equivalent image quality was obtained with the recent characterization and that performed 3 years prior. CONCLUSION: Model-based trajectory prediction enables high-quality image reconstruction for technically challenging sequences such as single-shot EPI and spiral imaging. It thus holds great promise for fast structural imaging and advanced neuroimaging techniques, including functional MRI, diffusion tensor imaging, and arterial spin labeling. The method can be based on a one-time system characterization as demonstrated by successful use of 3-year-old calibration data. Magn Reson Med 76:45-58, 2016. © 2015 Wiley Periodicals, Inc.

A field camera for MR sequence monitoring and system analysis.

PURPOSE: MR image formation and interpretation relies on highly accurate dynamic magnetic fields of high fidelity. A range of mechanisms still limit magnetic field fidelity, including magnet drifts, eddy currents, and finite linearity and stability of power amplifiers used to drive gradient and shim coils. Addressing remaining errors by means of hardware, sequence, or signal processing optimizations, calls for immediate observation by magnetic field monitoring. The present work presents a stand-alone monitoring system delivering insight into such field imperfections for MR sequence and system analysis. METHODS: A flexible NMR field probe-based stand-alone monitoring system, built on a software-defined-radio approach, is introduced and used to sense field dynamics up to third-order in space in a selection of situations with different time scales. RESULTS: Highly sensitive trajectories are measured and successfully used for image reconstruction. Further field perturbations due to mechanical oscillations and thermal field drifts following demanding gradient use and external interferences are studied. CONCLUSION: A flexible and versatile monitoring system is presented, delivering camera-like access to otherwise hardly accessible field dynamics with nanotesla resolution. Its stand-alone nature enables field analysis even during unknown MR system states.

Monitoring, analysis, and correction of magnetic field fluctuations in echo planar imaging time series.

PURPOSE: To assess the utility of concurrent magnetic field monitoring for observing and correcting for variations in k-space trajectories and global background fields that occur in single-shot echo planar imaging (EPI) time series as typically used in functional MRI (fMRI). METHODS: Field monitoring was performed using an array of NMR field probes operated concurrently with series of single-shot EPI acquisitions from a static phantom. The observed fluctuations in field evolution were analyzed in terms of their temporal and spatial behavior at the field level as well as at the level of reconstructed image series. The potential to correct for such fluctuations was assessed by accounting for them upon image reconstruction. An indication of the number and relative magnitude of underlying effects was obtained via principal component analysis. RESULTS: Trajectory and global field variations were found to induce substantial image fluctuations. Global field fluctuations induced standard deviations in image intensity up to 31%. Fluctuations in the trajectory induced ghosting artifacts with standard deviations up to 2%. Concurrent magnetic field monitoring reduced the fluctuations in the EPI time series to a maximum of 1.2%. CONCLUSION: Concurrent magnetic field monitoring holds the potential to improve the net sensitivity of fMRI by reducing signal fluctuations unrelated to brain activity.

Real-time feedback for spatiotemporal field stabilization in MR systems.

PURPOSE: MR imaging and spectroscopy require a highly stable, uniform background field. The field stability is typically limited by hardware imperfections, external perturbations, or field fluctuations of physiological origin. The purpose of the present work is to address these issues by introducing spatiotemporal field stabilization based on real-time sensing and feedback control. METHODS: An array of NMR field probes is used to sense the field evolution in a whole-body MR system concurrently with regular system operation. The field observations serve as inputs to a proportional-integral controller that governs correction currents in gradient and higher-order shim coils such as to keep the field stable in a volume of interest. RESULTS: The feedback system was successfully set up, currently reaching a minimum latency of 20 ms. Its utility is first demonstrated by countering thermal field drift during an EPI protocol. It is then used to address respiratory field fluctuations in a T2 *-weighted brain exam, resulting in substantially improved image quality. CONCLUSION: Feedback field control is an effective means of eliminating dynamic field distortions in MR systems. Third-order spatial control at an update time of 100 ms has proven sufficient to largely eliminate thermal and breathing effects in brain imaging at 7 Tesla.

Retrospective correction of physiological field fluctuations in high-field brain MRI using concurrent field monitoring.

PURPOSE: Magnetic field fluctuations caused by subject motion, such as breathing or limb motion, can degrade image quality in brain MRI, especially at high field strengths. The purpose of this study was to investigate the feasibility of retrospectively correcting for such physiological field perturbations based on concurrent field monitoring. THEORY AND METHODS: High-resolution T2*-weighted gradient-echo images of the brain were acquired at 7T with subjects performing different breathing and hand movement patterns. Field monitoring with a set of (19) F NMR probes distributed around the head was performed in two variants: concurrently with imaging or as a single field measurement per readout. The measured field fluctuations were then accounted for in the image reconstruction. RESULTS: Significant field fluctuations due to motion were observed in all subjects, resulting in severe artifacts in uncorrected images. The artifacts were largely removed by reconstruction based on field monitoring. Accounting for field perturbations up to the 1st spatial order was generally sufficient to recover good image quality. CONCLUSIONS: It has been demonstrated that artifacts due to physiologically induced dynamic field perturbations can be greatly reduced by retrospective image correction based on field monitoring. The necessity to perform such correction is greatest at high fields and for field-sensitive techniques such as T2*-weighted imaging.

Diffusion MRI with concurrent magnetic field monitoring.

PURPOSE: Diffusion MRI is compromised by unknown field perturbation during image encoding. The purpose of this study was to address this problem using the recently described approach of concurrent magnetic field monitoring. METHODS: Magnetic field dynamics were monitored during the echo planar imaging readout of a common diffusion-weighted MRI sequence using an integrated magnetic field camera setup. The image encoding including encoding changes over the duration of entire scans were quantified and analyzed. Field perturbations were corrected by accounting for them in generalized image reconstruction. The impact on image quality along with geometrical congruence among different diffusion-weighted images was assessed both qualitatively and quantitatively. RESULTS: The most significant field perturbations were found to be related to higher-order eddy currents from diffusion-weighting gradients and B0 field drift as well as gradual changes of short-term eddy current behavior and mechanical oscillations during the scan. All artifacts relating to dynamic field perturbations were eliminated by incorporating the measured encoding in image reconstruction. CONCLUSION: Concurrent field monitoring combined with generalized reconstruction enhances depiction fidelity in diffusion imaging. In addition to artifact reduction, it improves geometric congruence and thus facilitates image combination for quantitative diffusion analysis.

Matched-filter acquisition for BOLD fMRI.

We introduce matched-filter fMRI, which improves BOLD (blood oxygen level dependent) sensitivity by variable-density image acquisition tailored to subsequent image smoothing. Image smoothing is an established post-processing technique used in the vast majority of fMRI studies. Here we show that the signal-to-noise ratio of the resulting smoothed data can be substantially increased by acquisition weighting with a weighting function that matches the k-space filter imposed by the smoothing operation. We derive the theoretical SNR advantage of this strategy and propose a practical implementation of 2D echo-planar acquisition matched to common Gaussian smoothing. To reliably perform the involved variable-speed trajectories, concurrent magnetic field monitoring with NMR probes is used. Using this technique, phantom and in vivo measurements confirm reliable SNR improvement in the order of 30% in a "resting-state" condition and prove robust in different regimes of physiological noise. Furthermore, a preliminary task-based visual fMRI experiment equally suggests a consistent BOLD sensitivity increase in terms of statistical sensitivity (average t-value increase of about 35%). In summary, our study suggests that matched-filter acquisition is an effective means of improving BOLD SNR in studies that rely on image smoothing at the post-processing level.