Image reconstruction in SNR units: a general method for SNR measurement.

The method for phased array image reconstruction of uniform noise images may be used in conjunction with proper image scaling as a means of reconstructing images directly in SNR units. This facilitates accurate and precise SNR measurement on a per pixel basis. This method is applicable to root-sum-of-squares magnitude combining, B(1)-weighted combining, and parallel imaging such as SENSE. A procedure for image reconstruction and scaling is presented, and the method for SNR measurement is validated with phantom data. Alternative methods that rely on noise only regions are not appropriate for parallel imaging where the noise level is highly variable across the field-of-view. The purpose of this article is to provide a nuts and bolts procedure for calculating scale factors used for reconstructing images directly in SNR units. The procedure includes scaling for noise equivalent bandwidth of digital receivers, FFTs and associated window functions (raw data filters), and array combining.     

High temporal resolution functional MRI using parallel echo volumar imaging

PURPOSE: To combine parallel imaging with 3D single-shot acquisition (echo volumar imaging, EVI) in order to acquire high temporal resolution volumar functional MRI (fMRI) data. MATERIALS AND METHODS: An improved EVI sequence was associated with parallel acquisition and field of view reduction in order to acquire a large brain volume in 200 msec. Temporal stability and functional sensitivity were increased through optimization of all imaging parameters and Tikhonov regularization of parallel reconstruction. Two human volunteers were scanned with parallel EVI in a 1.5T whole-body MR system, while submitted to a slow event-related auditory paradigm. RESULTS: Thanks to parallel acquisition, the EVI volumes display a low level of geometric distortions and signal losses. After removal of low-frequency drifts and physiological artifacts, activations were detected in the temporal lobes of both volunteers and voxelwise hemodynamic response functions (HRF) could be computed. On these HRF different habituation behaviors in response to sentence repetition could be identified. CONCLUSION: This work demonstrates the feasibility of high temporal resolution 3D fMRI with parallel EVI. Combined with advanced estimation tools, this acquisition method should prove useful to measure neural activity timing differences or study the nonlinearities and nonstationarities of the BOLD response.     

Parallel MR imaging

Parallel imaging is a robust method for accelerating the acquisition of magnetic resonance imaging (MRI) data, and has made possible many new applications of MR imaging. Parallel imaging works by acquiring a reduced amount of k-space data with an array of receiver coils. These undersampled data can be acquired more quickly, but the undersampling leads to aliased images. One of several parallel imaging algorithms can then be used to reconstruct artifact-free images from either the aliased images (SENSE-type reconstruction) or from the undersampled data (GRAPPA-type reconstruction). The advantages of parallel imaging in a clinical setting include faster image acquisition, which can be used, for instance, to shorten breath-hold times resulting in fewer motion-corrupted examinations. In this article the basic concepts behind parallel imaging are introduced. The relationship between undersampling and aliasing is discussed and two commonly used parallel imaging methods, SENSE and GRAPPA, are explained in detail. Examples of artifacts arising from parallel imaging are shown and ways to detect and mitigate these artifacts are described. Finally, several current applications of parallel imaging are presented and recent advancements and promising research in parallel imaging are briefly reviewed.     

Design of a SENSE-optimized high-sensitivity MRI receive coil for brain imaging.

An 8-channel receive-only detector array was developed for SENSE MRI of human brain. The coil geometry was based on a gapped element design and used ultra-high impedance preamplifiers for mutual decoupling of the elements. Computer simulations of the electric and magnetic fields showed that excellent signal-to-noise ratio (SNR) and SENSE performance could be achieved by placing the coil elements close to the head and maintaining a substantial gap between the elements. Measurements with a 1.5 T prototype coil showed a 2.7-fold improvement of the SNR averaged over the brain compared to a conventional quadrature birdcage receive coil and an average geometrical noise amplification factor (g-value) of 1.06 and 1.38 for rate-2 and rate-3 SENSE, respectively.     

Parallel imaging performance as a function of field strength--an experimental investigation using electrodynamic scaling.

In this work, the dependence of parallel MRI performance on main magnetic field strength is experimentally investigated. Using the general framework of electrodynamic scaling, the B0-dependent behavior of the relevant radiofrequency fields at a single physical field strength of 7 T is studied. In the chosen implementation this is accomplished by adjusting the permittivity and conductivity of a homogeneous spherical phantom. With different mixing ratios of decane, ethanol, purified water, N-methylformamide, and sodium chloride, field strengths in the range of 1.5 to 11.5 T are mimicked. Based on sensitivity maps of an eight-coil receiver array, the field-dependent performance of parallel imaging is assessed in terms of the geometry factor g, which reflects noise enhancement in parallel imaging reconstruction. At low field strengths the SNR penalty was nearly independent of B0 and favorably low for 1D reduction factors up to between 3 and 4. At higher field strengths the transition between favorable and prohibitive parallel imaging conditions was found to shift toward higher feasible reduction factors. These findings are in good agreement with previous theoretical predictions. From this agreement it is concluded that parallel MRI at high B0 benefits specifically from onsetting far-field behavior of the involved radiofrequency fields.     

Functional MRI using regularized parallel imaging acquisition.

Parallel MRI techniques reconstruct full-FOV images from undersampled k-space data by using the uncorrelated information from RF array coil elements. One disadvantage of parallel MRI is that the image signal-to-noise ratio (SNR) is degraded because of the reduced data samples and the spatially correlated nature of multiple RF receivers. Regularization has been proposed to mitigate the SNR loss originating due to the latter reason. Since it is necessary to utilize static prior to regularization, the dynamic contrast-to-noise ratio (CNR) in parallel MRI will be affected. In this paper we investigate the CNR of regularized sensitivity encoding (SENSE) acquisitions. We propose to implement regularized parallel MRI acquisitions in functional MRI (fMRI) experiments by incorporating the prior from combined segmented echo-planar imaging (EPI) acquisition into SENSE reconstructions. We investigated the impact of regularization on the CNR by performing parametric simulations at various BOLD contrasts, acceleration rates, and sizes of the active brain areas. As quantified by receiver operating characteristic (ROC) analysis, the simulations suggest that the detection power of SENSE fMRI can be improved by regularized reconstructions, compared to unregularized reconstructions. Human motor and visual fMRI data acquired at different field strengths and array coils also demonstrate that regularized SENSE improves the detection of functionally active brain regions.     

Field-of-view limitations in parallel imaging.

Parallel imaging is one of the most promising developments in recent years for the acceleration of MR acquisitions. One area of practical importance where different parallel imaging methods perform differently is the manner in which they deal with aliasing in the full-FOV reconstructed image. It has been reported that sensitivity encoding (SENSE) reconstruction fails whenever the reconstructed FOV is smaller than the object being imaged. On the other hand, generalized autocalibrating partially parallel acquisition (GRAPPA) has been used successfully to reconstruct images with aliasing in the reconstructed FOV, as in conventional imaging. The disparate behavior of these methods can be easily demonstrated by a few simple illustrative examples. Additional in vivo examples using GRAPPA and modified SENSE (mSENSE) make this distinction clear. These experiments demonstrate that SENSE fails to reconstruct correct images when coil sensitivity maps are used that do not automatically account for the object size and therefore the aliasing in the reconstructed images. However, with the use of aliased high-resolution coil sensitivity maps, accurate SENSE reconstructions can be generated. On the other hand, GRAPPA produces images with an aliasing appearance that is exactly as would be expected from normal nonaccelerated acquisitions. An understanding of these effects could potentially lead to fewer operator-dependent errors, as well as a better understanding of the differences between the underlying reconstruction processes.     

Sensitivity encoding as a means of enhancing the SNR efficiency in steady-state MRI.

Sensitivity encoding (SENSE) with a receiver coil array is typically used as a means of reducing the scan time in MRI. The speed benefit usually comes at some expense in terms of the signal-to-noise ratio (SNR) efficiency, which has been notorious as the main downside of SENSE and parallel MRI in general. In this work it is shown that in steady-state gradient-echo imaging the parallel approach may as well be used to increase the SNR efficiency. The basic idea is to balance reduced phase encoding by increasing the repetition time. In this fashion both the acquisition duty cycle and the steady-state magnetization can be enhanced, resulting in considerable net gains in SNR yield. It is argued that the reduction factor in parallel imaging is essentially an additional degree of freedom in optimizing the SNR. The optimal SENSE factor depends on scan, tissue, and hardware parameters, assuming values up to 3.0 and higher. The achievable SNR benefit also depends on the spoiling regime and is most pronounced for RF-spoiled techniques. The proposed mechanism is demonstrated by simulations and phantom experiments, as well as by contrast-enhanced angiography in vivo, achieving an approximate doubling of the SNR efficiency.     

Effects of inductive coupling on parallel MR image reconstructions.

Theoretical arguments and experimental results are presented that characterize the impact of inductive coupling on the performance of parallel MRI reconstructions. A simple model of MR signal and noise reception suggests that the intrinsic amount of spatial information available from a given coil array is unchanged in the presence of inductive coupling, as long as the sample remains the dominant source of noise for the coupled array. Any loss of distinctness in the measured coil sensitivities is compensated by information stored in the measured noise correlations. Adjustments to the theory are described to account for preamplifier noise contributions. Results are presented from an experimental system in which preamplifier input impedances are systematically adjusted in order to vary the level of coupling between array elements. Parallel image reconstructions using an array with four different levels of coupling and an acceleration factor up to six show average SNR changes of -7.6% to +7.5%. The modest changes in overall SNR are accompanied by similarly small changes in g-factor. These initial results suggest that moderate amounts of inductive coupling should not have a prohibitive effect on the use of a given coil array for parallel MRI.     

Self-calibrating parallel imaging with automatic coil sensitivity extraction.

Calibration of the spatial sensitivity functions of coil arrays is a crucial element in parallel magnetic resonance imaging (PMRI). The most common approach has been to measure coil sensitivities directly using one or more low-resolution images acquired before or after accelerated data acquisition. However, since it is difficult to ensure that the patient and coil array will be in exactly the same positions during both calibration scans and accelerated imaging, this approach can introduce sensitivity miscalibration errors into PMRI reconstructions. This work shows that it is possible to extract sensitivity calibration images directly from a fully sampled central region of a variable-density k-space acquisition. These images have all the features of traditional PMRI sensitivity calibrations and therefore may be used for any PMRI reconstruction technique without modification. Because these calibration data are acquired simultaneously with the data to be reconstructed, errors due to sensitivity miscalibration are eliminated. In vivo implementations of self-calibrating parallel imaging using a flexible coil array are demonstrated in abdominal imaging and in real-time cardiac imaging studies.     

Influence of SENSE on image properties in high-resolution single-shot echo-planar DTI.

Limited spatial resolution is a key obstacle to the study of brain white matter structure with diffusion tensor imaging (DTI). In its frequent implementation with single-excitation spin-echo echo-planar sequences, DTI's ability to resolve small structures is strongly restricted by T2 and T2* decay, B0 inhomogeneity, and limited signal-to-noise ratio (SNR). In this work the influence of sensitivity encoding (SENSE) on diffusion-weighted (DW) image properties is investigated. Computer simulations showed that the PSF becomes narrower with increasing SENSE reduction factors, R, enhancing the intrinsic resolution. After a brief theoretical discussion, we describe the estimation of SNR on a pixel-by-pixel basis as a function of R. The mean image SNR behavior is manifold: SENSE is capable of increasing SNR efficiency by reducing the echo time (TE). Each SNR(R) curve reveals a maximum that depends on the amount of partial Fourier encoding used. The overall best SNR efficiency for an eight-element head coil array and a b-factor of 1000 s/mm2 is achieved at R = 2.1 and partial Fourier encoding of 60%. In vivo tensor maps of volunteers and a patient, with an in-plane resolution of 0.78 x 0.78 mm2, are also presented to demonstrate the practical implementation of the parallel approach.     

Ghost artifact removal using a parallel imaging approach.

Parallel imaging techniques, which use several receive coils simultaneously, have been shown to enable a significant scan time reduction by subsampling k-space. Nevertheless, the data acquired with multiple coils in parallel exhibit some redundancy if the number of receive coils exceeds the subsampling factor. This redundancy leads to an overdetermination of the reconstruction problem, which is generally used to optimize the signal-to-noise ratio (SNR). However, it can yield further information about the quality of the reconstructed image, and can thus be used to identify and correct image artifacts. While some known approaches try to solve this problem in k-space, this study addresses it in the spatial domain and uses a modified SENSE reconstruction to reduce or completely remove ghost-type artifacts arising from processes such as motion or flow during data acquisition. Phantom and in vivo studies show significant improvements in image quality after correction, and serve as a basis for the discussion of the performance and limitations of this new approach.     

Continuously moving table SENSE imaging.

A combination of continuously moving table imaging and parallel imaging based on sensitivity encoding (SENSE) is presented. One specific geometry is considered, where the receiver array is fixed to the MR magnet and does not move with the table, which allows for head-to-toe imaging with a small total number of coils. Sensitivity maps are defined for the enlarged virtual field of view and are composed according to the k-space sampling scheme such that established parallel reconstruction techniques are applicable to good approximation. In vivo experiments show the feasibility of this approach, and simulations determine the application range. Three-dimensional head-to-toe imaging of volunteers is performed in 77 s with a SENSE reduction factor of 2 in a virtual field of view of 1800 x 460 x 100 mm(3).     

Controlled aliasing in parallel imaging results in higher acceleration (CAIPIRINHA) for multi-slice imaging.

In all current parallel imaging techniques, aliasing artifacts resulting from an undersampled acquisition are removed by means of a specialized image reconstruction algorithm. In this study a new approach termed "controlled aliasing in parallel imaging results in higher acceleration" (CAIPIRINHA) is presented. This technique modifies the appearance of aliasing artifacts during the acquisition to improve the subsequent parallel image reconstruction procedure. This new parallel multi-slice technique is more efficient compared to other multi-slice parallel imaging concepts that use only a pure postprocessing approach. In this new approach, multiple slices of arbitrary thickness and distance are excited simultaneously with the use of multi-band radiofrequency (RF) pulses similar to Hadamard pulses. These data are then undersampled, yielding superimposed slices that appear shifted with respect to each other. The shift of the aliased slices is controlled by modulating the phase of the individual slices in the multi-band excitation pulse from echo to echo. We show that the reconstruction quality of the aliased slices is better using this shift. This may potentially allow one to use higher acceleration factors than are used in techniques without this excitation scheme. Additionally, slices that have essentially the same coil sensitivity profiles can be separated with this technique.     

Accuracy and reproducibility in phase contrast imaging using SENSE.

The purpose of this study was to evaluate the accuracy and reproducibility of phase contrast imaging using the sensitivity encoding (SENSE) method at different reduction factors. Analytical expressions were derived that state how reproducibility is influenced for velocity and flow measurements. Computer simulations, and in vitro and in vivo studies were performed in order to validate these expressions and to assess how accuracy is affected when different reduction factors are applied. It was shown that reproducibility depends on the reduction and geometry factors. Since the geometry factor varies spatially, so does the reproducibility for phase contrast imaging. In areas with high geometry factors, the standard deviation (SD) may become so large that aliasing occurs. The accuracy of phase contrast imaging is not influenced directly when SENSE is used, but may be indirectly influenced due to high SDs of the measured phase that may subsequently cause aliasing. The current results show that it is possible to achieve accurate flow measurements even at high reduction factors. By taking the geometry factor into account, it may be possible to find areas where phase contrast imaging is accurate even at high reduction factors.     

Auto-calibrated parallel spiral imaging.

This work describes an auto-calibrated method for parallel imaging with spiral trajectory. The method is a k-space approach where an interpolation kernel, accounting for coil sensitivity factors, is derived from experimental data and used to interpolate the reduced data set in parallel imaging to estimate the missing k-space data. For the case of spiral imaging, this interpolation kernel is defined along radial directions so that missing spiral interleaves can be estimated directly from neighboring interleaves. This kernel is invariant along the radial direction but varies azimuthally. Therefore, the k-space is divided into angular sectors and sector-specific kernels are used. It is demonstrated experimentally that relatively few sectors are sufficient for accurate reconstruction, allowing for efficient implementation. The interpolation kernels can be derived either from a separate calibration scan or self-calibration data available with a dual-density spiral acquisition. The reconstruction method is implemented with two sampling strategies and experimentally demonstrated to be robust.     

Bayesian parallel imaging with edge-preserving priors.

Existing parallel MRI methods are limited by a fundamental trade-off in that suppressing noise introduces aliasing artifacts. Bayesian methods with an appropriately chosen image prior offer a promising alternative; however, previous methods with spatial priors assume that intensities vary smoothly over the entire image, resulting in blurred edges. Here we introduce an edge-preserving prior (EPP) that instead assumes that intensities are piecewise smooth, and propose a new approach to efficiently compute its Bayesian estimate. The estimation task is formulated as an optimization problem that requires a nonconvex objective function to be minimized in a space with thousands of dimensions. As a result, traditional continuous minimization methods cannot be applied. This optimization task is closely related to some problems in the field of computer vision for which discrete optimization methods have been developed in the last few years. We adapt these algorithms, which are based on graph cuts, to address our optimization problem. The results of several parallel imaging experiments on brain and torso regions performed under challenging conditions with high acceleration factors are shown and compared with the results of conventional sensitivity encoding (SENSE) methods. An empirical analysis indicates that the proposed method visually improves overall quality compared to conventional methods.     

Sensitivity-encoded (SENSE) proton echo-planar spectroscopic imaging (PEPSI) in the human brain.

Magnetic resonance spectroscopic imaging (MRSI) provides spatially resolved metabolite information that is invaluable for both neuroscience studies and clinical applications. However, lengthy data acquisition times, which are a result of time-consuming phase encoding, represent a major challenge for MRSI. Fast MRSI pulse sequences that use echo-planar readout gradients, such as proton echo-planar spectroscopic imaging (PEPSI), are capable of fast spectral-spatial encoding and thus enable acceleration of image acquisition times. Combining PEPSI with recent advances in parallel MRI utilizing RF coil arrays can further accelerate MRSI data acquisition. Here we investigate the feasibility of ultrafast spectroscopic imaging at high field (3T and 4T) by combining PEPSI with sensitivity-encoded (SENSE) MRI using eight-channel head coil arrays. We show that the acquisition of single-average SENSE-PEPSI data at a short TE (15 ms) can be accelerated to 32 s or less, depending on the field strength, to obtain metabolic images of choline (Cho), creatine (Cre), N-acetyl-aspartate (NAA), and J-coupled metabolites (e.g., glutamate (Glu) and inositol (Ino)) with acceptable spectral quality and localization. The experimentally measured reductions in signal-to-noise ratio (SNR) and Cramer-Rao lower bounds (CRLBs) of metabolite resonances were well explained by both the g-factor and reduced measurement times. Thus, this technology is a promising means of reducing the scan times of 3D acquisitions and time-resolved 2D measurements.     

Routine MR examination of the knee using parallel imaging

AIM: To demonstrate the feasibility and the clinical value of parallel imaging of the knee. METHODS: A pair of flex coils was used for the parallel imaging. Before optimization of the protocol, qualitative and quantitative analysis was performed by volunteer and phantom study. Two separate consecutive groups of 80 patients were examined with or without parallel imaging and compared with the arthroscopic results. RESULTS: Parallel imaging obtained a total of 164 images of six series in less than 20 min. Compared with conventional imaging, the parallel imaging appeared slightly noisy with an 83% relative signal-to-noise ratio. Demarcation of the cruciate ligaments and menisci were rated equivalent. For the patients study, the parallel imaging delineated ligament and meniscus pathologies with accuracies of over 93%. CONCLUSION: Parallel MR imaging of the knee enables significant reduction of scan time as well as preserving diagnostic results.