Sparsity-promoting regularization is useful for combining compressed sensing assumptions with parallel MRI for reducing scan time while preserving image quality. Variable splitting algorithms are the current state-of-the-art algorithms for SENSE-type MR image reconstruction with sparsity-promoting regularization. These methods are very general and have been observed to work with almost any regularizer; however, the tuning of associated convergence parameters is a commonly-cited hindrance in their adoption. Conversely, majorize-minimize algorithms based on a single Lipschitz constant have been observed to be slow in shift-variant applications such as SENSE-type MR image reconstruction since the associated Lipschitz constants are loose bounds for the shift-variant behavior. This paper bridges the gap between the Lipschitz constant and the shift-variant aspects of SENSE-type MR imaging by introducing majorizing matrices in the range of the regularizer matrix. The proposed majorize-minimize methods (called BARISTA) converge faster than state-of-the-art variable splitting algorithms when combined with momentum acceleration and adaptive momentum restarting. Furthermore, the tuning parameters associated with the proposed methods are unitless convergence tolerances that are easier to choose than the constraint penalty parameters required by variable splitting algorithms.
MR Image Reconstruction; Compressed Sensing; FISTA; Majorize-Minimize; Parallel MRI
This work presents a 3D tailored RF slab-select pulse for reducing the B1 field inhomogeneity at 3T and possibly higher applied field strengths. The small tip angle approximation is used to design a compensated 3D slice profile from a map of the B1 inhomogeneity. The SNR improvement and degree and artifact reduction were evaluated in NiCl2 doped phantom experiments. The method was also implemented in-vivo in the human brain where it was found to be effective in reducing inhomogeneities as large as 35% of the peak image magnitude in all three spatial directions in the brain. The main drawback of the technique is long pulse lengths.
A fast parallel excitation pulse design algorithm to select and to order Phase-Encoding (PE) locations (also known as “spokes”) of an Echo-Volumar excitation k-space trajectory considering B0 field inhomogeneity is presented. Recently, other groups have conducted research to choose optimal PE locations, but the potential benefit of considering B0 field inhomogeneity during PE location selection or their ordering has not been fully investigated. This paper introduces a novel fast greedy algorithm to determine PE locations and their order that takes into account the off-resonance effects. Computer simulations of the proposed algorithm for B1 field inhomogeneity correction demonstrate that it not only improves excitation accuracy but also provides an effective ordering of the PE locations.
parallel excitation; RF pulse design; PE location selection; Spoke selection; EV trajectory
Small-tip fast recovery (STFR) imaging is a new steady-state imaging sequence that is a potential alternative to balanced steady-state free precession (bSSFP). Under ideal imaging conditions, STFR may provide comparable signal-to-noise ratio (SNR) and image contrast as bSSFP, but without signal variations due to resonance offset. STFR relies on a tailored “tip-up”, or “fast recovery”, RF pulse to align the spins with the longitudinal axis after each data readout segment. The design of the tip-up pulse is based on the acquisition of a separate off-resonance (B0) map. Unfortunately, the design of fast (a few ms) slice- or slab-selective RF pulses that accurately tailor the excitation pattern to the local B0 inhomogeneity over the entire imaging volume remains a challenging and unsolved problem. We introduce a novel implementation of STFR imaging based on non-slice-selective tip-up pulses, which simplifies the RF design problem significantly. Out-of-slice magnetization pathways are suppressed using RF-spoiling. Brain images obtained with this technique show excellent gray/white matter contrast, and point to the possibility of rapid steady-state T2/T1-weighted imaging with intrinsic suppression of cerebrospinal fluid, through-plane vessel signal, and off-resonance artifacts. In the future we expect STFR imaging to benefit significantly from parallel excitation hardware and high-order gradient shim systems.
steady state free precession; banding artifacts; inflow suppression; CSF suppression
To develop an approach for computer-aided detection (CAD) of small brain metastases in post-Gd T1-weighted magnetic resonance imaging (MRI).
A set of unevenly spaced 3D spherical shell templates was optimized to localize brain metastatic lesions by cross-correlation analysis with MRI. Theoretical and simulation analyses of effects of lesion size and shape heterogeneity were performed to optimize the number and size of the templates and the cross-correlation thresholds. Also, effects of image factors of noise and intensity variation on the performance of the CAD system were investigated. A nodule enhancement strategy to improve sensitivity of the system and a set of criteria based upon the size, shape and brightness of lesions were used to reduce false positives. An optimal set of parameters from the FROC curves was selected from a training dataset, and then the system was evaluated on a testing dataset including 186 lesions from 2753 MRI slices. Reading results from two radiologists are also included.
Overall, a 93.5% sensitivity with 0.024 of intra-cranial false positive rate (IC-FPR) was achieved in the testing dataset. Our investigation indicated that nodule enhancement was very effective in improving both sensitivity and specificity. The size and shape criteria reduced the IC-FPR from 0.075 to 0.021, and the brightness criterion decreases the extra-cranial FPR from 0.477 to 0.083 in the training dataset. Readings from the two radiologists had sensitivities of 60% and 67% in the training dataset and 70% and 80% in the testing dataset for the metastatic lesions <5 mm in diameter.
Our proposed CAD system has high sensitivity and fairly low FPR for detection of the small brain metastatic lesions in MRI compared to the previous work and readings of neuroradiologists. The potential of this method for assisting clinical decision- making warrants further evaluation and improvements.
Brain Metastasis; Computer-Aided Detection; Template Matching; MRI
Compressed sensing (CS) has been used for accelerating magnetic resonance imaging (MRI) acquisitions, but its use in applications with rapid spatial phase variations is challenging, e.g., proton resonance frequency shift (PRF-shift) thermometry and velocity mapping. Previously, an iterative MRI reconstruction with separate magnitude and phase regularization was proposed for applications where magnitude and phase maps are both of interest, but it requires fully sampled data and unwrapped phase maps. In this paper, CS is combined into this framework to reconstruct magnitude and phase images accurately from undersampled data. Moreover, new phase regularization terms are proposed to accommodate phase wrapping and to reconstruct images with encoded phase variations, e.g., PRF-shift thermometry and velocity mapping. The proposed method is demonstrated with simulated thermometry data and in-vivo velocity mapping data and compared to conventional phase corrected CS.
Compressed sensing; regularization; image reconstruction; magnetic resonance imaging
Cushing syndrome (CS) is the classic condition of cortisol dysregulation, and cortisol dysregulation is the prototypic finding in Major Depressive Disorder (MDD). We hypothesized that subjects with active CS would show dysfunction in frontal and limbic structures relevant to affective networks, and also manifest poorer facial affect identification accuracy, a finding reported in MDD.Twenty-one patients with confirmed CS (20 ACTH-dependent and 1 ACTH-independent) were compared to 21 healthy controlsubjects. Identification of affective facial expressions (Facial Emotion Perception Test) was conducted in a 3 Tesla GE fMRI scanner using BOLD fMRI signal. The impact of disease (illness duration, current hormone elevation and degree of disruption of circadian rhythm), performance, and comorbid conditions secondary to hypercortisolemia were evaluated.CS patients made more errors in categorizing facial expressions and had less activation in left anterior superior temporal gyrus, a region important in emotion processing. CS patients showed higher activation in frontal, medial, and subcortical regions relative to controls. Two regions of elevated activation in CS, left middle frontal and lateral posterior/pulvinar areas, were positively correlated with accuracy in emotion identification in the CS group, reflecting compensatory recruitment. In addition, within the CSgroup, greater activation in left dorsal anterior cingulatewas related to greater severity of hormone dysregulation. In conclusion, cortisol dysregulation in CS patients is associated with problems in accuracy of affective discrimination and altered activation of brain structures relevant to emotion perception, processing and regulation, similar to the performance decrements and brain regions shown to be dysfunctional in MDD.
HPA; cortisol; ACTH; emotion; affect; fMRI; Cushings
The ability to distinguish between high and low levels of task engagement in the real world is important for detecting and preventing performance decrements during safety-critical operational tasks. We therefore investigated whether functional Near Infrared Spectroscopy (fNIRS), a portable brain neuroimaging technique, can be used to distinguish between high and low levels of task engagement during the performance of a selective attention task. A group of participants performed the multi-source interference task (MSIT) while we recorded brain activity with fNIRS from two brain regions. One was a key region of the “task-positive” network, which is associated with relatively high levels of task engagement. The second was a key region of the “task-negative” network, which is associated with relatively low levels of task engagement (e.g., resting and not performing a task). Using activity in these regions as inputs to a multivariate pattern classifier, we were able to predict above chance levels whether participants were engaged in performing the MSIT or resting. We were also able to replicate prior findings from functional magnetic resonance imaging (fMRI) indicating that activity in task-positive and task-negative regions is negatively correlated during task performance. Finally, data from a companion fMRI study verified our assumptions about the sources of brain activity in the fNIRS experiment and established an upper bound on classification accuracy in our task. Together, our findings suggest that fNIRS could prove quite useful for monitoring cognitive state in real-world settings.
near infra-red spectroscopy; attention; default mode network; classification; human performance
This paper presents a small-flip-angle, three-dimensional tailored RF pulse that excites thin slices with an adjustable quadratic in-plane spatial variation. The quadratic spatial variation helps to compensate for the loss in image uniformity using a volume coil at 3T due to the wavelike properties of the RF field. The pulse is based on a novel “fast-kz” design that uses a series of slice-select sub-pulses along kz and phase encoding “blips” along kx-ky. The method is demonstrated by acquiring a series of 5 mm thick T2-weighted images of the human brain at 3T using pulses 4.8 ms in length with a 45° flip angle.
MRI; Human brain; B1 inhomogeneity; high magnetic fields; tailored RF pulses
A multi-shot three-dimensional slice-select tailored RF pulse method is presented for the excitation of slice profiles with arbitrary resolution. This method is derived from the linearity of the small tip angle approximation, allowing for the decomposition of small tip angle tailored RF pulses into separate excitations. The final image is created by complex summation of the images acquired from the individual excitations. This technique overcomes the limitation of requiring long pulse to excite thin slices with adequate resolution. This has implications in applications including T2*-weighted functional MRI in brain regions corrupted by intravoxel dephasing artifacts due to susceptibility variations. Simulations, phantom experiments, and human brain images are presented. It is demonstrated that at most four shots of 40 ms pulse length are needed to excite a 5 mm thick slice in the brain with reduced susceptibility artifacts at 3T.
Functional MRI; susceptibility artifacts; 3D tailored RF pulses; spiral imaging
An extension of the “UNaliasing by Fourier encoding the Overlaps using the temporaL Dimension” (UNFOLD) method to excitation domain (XUNFOLD) is presented that reduces the effective length of small tip angle tailored RF pulses. Multi-shot tailored RF pulses were designed such that each shot could be alternated every TR in a dynamic-imaging application, producing a time series of images with periodic aliased slice profiles. The XUNFOLD method removes the slice-profile aliasing from the complex four-dimensional (xyzt) image data by filtering in the temporal frequency dimension identical to the UNFOLD method. The XUNFOLD technique is demonstrated using 3D tailored RF pulses designed for reducing intravoxel dephasing due to magnetic susceptibility variations in T2*-weighted brain imaging at 3T. The use of XUNFOLD allows for a single-shot implementation of the 3D tailored RF method for reducing susceptibility artifacts.
A three-dimensional tailored RF pulse method for reducing intravoxel dephasing artifacts in T2* weighted functional MRI is presented. A stack of spirals k-space trajectory is employed to excite a disk of magnetization for small tip angles. Smaller disks with a linear through-plane phase are inserted into the disk to locally refocus regions which are normally dephased due to susceptibility variations. Numerical simulations and imaging experiments which use the tailored RF pulses are presented. Limitations of the method and improvements are also discussed.
Functional MRI; susceptibility artifacts; tailored RF pulses
Parallel transmitter techniques are a promising approach for reducing transmitter B1 inhomogeneity due to the potential for adjusting the spatial excitation profile with independent RF pulses. These techniques may be further improved with transmit sensitivity encoding (SENSE) methods because the sensitivity information in pulse design provides an excitation that is inherently compensated for transmitter B1 inhomogeneity. This paper presents a proof of this concept using transmit SENSE three-dimensional tailored RF pulses designed for small flip angles. An eight-channel receiver coil was used to mimic parallel transmission for brain imaging at 3T. The transmit SENSE pulses were based on the fast-kz design and produced 5 mm thick slices at a flip angle of 30° with only a 4.3 ms pulse length. It was found that the transmit SENSE pulses produced more homogenous images than those obtained from the complex sum of images from all receivers excited with a standard RF pulse.
Brain imaging; B1 inhomogeneity; Transmit SENSE; parallel transmitters; tailored RF pulses
A variable density spiral method is presented for reducing three-dimensional tailored RF pulse duration. Pulse length reductions of 24-32% are possible with no loss in excitation resolution at the expense of a 3-5% increase in slice profile non-uniformity. The method is demonstrated using simulations, phantom experiments, and T2*-weighted images of brain regions with susceptibility induced intravoxel dephasing. Four 13.2 ms shots were needed to excite a 5 mm thick slice with reduced susceptibility artifact in the sinus region at 3T.
Variable density spirals; tailored RF pulses; susceptibility artifacts; functional MRI
Inhibitory control or regulatory difficulties have been explored in major depressive disorder (MDD) but typically in the context of affectively salient information. Inhibitory control is addressed specifically by using a task devoid of affectively-laden stimuli, to disentangle the effects of altered affect and altered inhibitory processes in MDD.
Twenty MDD and 22 control volunteer participants matched by age and gender completed a contextual inhibitory control task, the Parametric Go/No-go (PGNG) task during functional magnetic resonance imaging. The PGNG includes three levels of difficulty, a typical continuous performance task and two progressively more difficult versions including Go/No-go hit and rejection trials. After this test, 15 of 20 MDD patients completed a full 10-week treatment with s-citalopram.
There was a significant interaction among response time (control subjects better), hits (control subjects better), and rejections (patients better). The MDD participants had greater activation compared with the control group in frontal and anterior temporal areas during correct rejections (inhibition). Activation during successful inhibitory events in bilateral inferior frontal and left amygdala, insula, and nucleus accumbens and during unsuccessful inhibition (commission errors) in rostral anterior cingulate predicted post-treatment improvement in depression symptoms.
The imaging findings suggest that in MDD subjects, greater neural activation in frontal, limbic, and temporal regions during correct rejection of lures is necessary to achieve behavioral performance equivalent to control subjects. Greater activation in similar regions was further predictive of better treatment response in MDD.
Depression; executive functioning; fMRI; imaging; inhibitory control; mood disorders; SSRIs; treatment response
T2∗-weighted functional MR images suffer from signal loss artifacts caused by the magnetic susceptibility differences between air cavities and brain tissues. We propose a novel spectral-spatial pulse design that is slice-selective and capable of mitigating the signal loss. The two-dimensional spectral-spatial pulses create precompensatory phase variations that counteract through-plane dephasing, relying on the assumption that resonance frequency offset and through-plane field gradient are spatially correlated. The pulses can be precomputed prior to functional MRI experiments and used repeatedly for different slices in different subjects. Experiments with human subjects showed that the pulses were effective in slice selection and loss mitigation at different brain regions.
In fast MR imaging with long readout times, such as echo-planar imaging (EPI) and spiral scans, it is important to correct for the effects of field inhomogeneity to reduce image distortion and blurring. Such corrections require an accurate field map, a map of the off-resonance frequency at each voxel. Standard field map estimation methods yield noisy field maps, particularly in image regions with low spin density. This paper, describes regularized methods for field map estimation from two or more MR scans having different echo times. These methods exploit the fact that field maps are generally smooth functions. The methods use algorithms that decrease monotonically a regularized least-squares cost function, even though the problem is highly nonlinear. Results show that the proposed regularized methods significantly improve the quality of field map estimates relative to conventional unregularized methods.
Large-tip-angle multidimensional RF pulse design is a difficult problem, due to the nonlinear response of magnetization to applied RF at large tip-angles. In parallel excitation, multidimensional RF pulse design is further complicated by the possibility for transmit field patterns to change between subjects, requiring pulses to be designed rapidly while a subject lies in the scanner. To accelerate pulse design, we introduce a fast version of the optimal control method for large-tip-angle parallel excitation. The new method is based on a novel approach to analytically linearizing the Bloch equation about a large-tip-angle RF pulse, which results in an approximate linear model for the perturbations created by adding a small-tip-angle pulse to a large-tip-angle pulse. The linear model can be evaluated rapidly using non-uniform fast Fourier transforms, and we apply it iteratively to produce a sequence of pulse updates that improve excitation accuracy. We achieve drastic reductions in design time and memory requirements compared to conventional optimal control, while producing pulses of similar accuracy. The new method can also compensate for non-idealities such as main field inhomogeneties.
MRI; RF pulse design; large-tip-angle RF pulse design; multidimensional excitation; parallel excitation