Evaluation of highly accelerated simultaneous multi-slice EPI for fMRI.

Echo planar imaging (EPI) is the MRI technique that is most widely used for blood oxygen level-dependent (BOLD) functional MRI (fMRI). Recent advances in EPI speed have been made possible with simultaneous multi-slice (SMS) methods which combine acceleration factors M from multiband (MB) radiofrequency pulses and S from simultaneous image refocusing (SIR) to acquire a total of N=S×M images in one echo train, providing up to N times speed-up in total acquisition time over conventional EPI. We evaluated accelerations as high as N=48 using different combinations of S and M which allow for whole brain imaging in as little as 100ms at 3T with a 32 channel head coil. The various combinations of acceleration parameters were evaluated by tSNR as well as BOLD contrast-to-noise ratio (CNR) and information content from checkerboard and movie clips in fMRI experiments. We found that at low acceleration factors (N≤6), setting S=1 and varying M alone yielded the best results in all evaluation metrics, while at acceleration N=8 the results were mixed using both S=1 and S=2 sequences. At higher acceleration factors (N>8), using S=2 yielded maximal BOLD CNR and information content as measured by classification of movie clip frames. Importantly, we found significantly greater BOLD information content using relatively fast TRs in the range of 300ms-600ms compared to a TR of 2s, suggesting that faster TRs capture more information per unit time in task based fMRI.

Single-shot 3D GRASE with cylindrical k-space trajectories.

Development of GRASE (gradient- and spin-echo) pulse sequences for single-shot 3D imaging has been motivated by physiologic studies of the brain. The duration of echo-planar imaging (EPI) subsequences between RF refocusing pulses in the GRASE sequence is determinant of image distortions and susceptibility artifacts. To reduce these artifacts the regular Cartesian trajectory is modified to a circular trajectory in 2D and a cylindrical trajectory in 3D for reduced echo train time. Incorporation of "fly-back" trajectories lengthened the time of the subsequences and proportionally increased susceptibility artifact but the unipolar readout gradients eliminate all ghost artifacts. The modified cylindrical trajectory reduced susceptibility artifact and distortion artifact while raising the signal-to-noise ratio in both phantom and human brain images.

Changes in size and magnetic resonance signal intensity of the cerebral CSF spaces during the cardiac cycle as studied by gated, high-resolution magnetic resonance imaging.

In 1966, du Boulay demonstrated the pulsatile nature of CSF flow in the cerebral aqueduct by using air cineventriculography, which disturbs normal CSF dynamics by replacing part of the incompressible CSF with air. To investigate this phenomenon noninvasively, 35 normal volunteers were studied using high-resolution, cardiac-gated MR imaging. Specifically, we wished to document changes in size and configuration of the CSF spaces and the incidence and magnitude of signal loss (an indication of CSF motion) in these spaces as they related to time in the cardiac cycle. Changes in size and configuration were measurable in the third ventricle only (size increased during systole in seven of the 35 volunteers). Except for the lateral ventricles, some loss in signal intensity was seen in all CSF spaces at least during systole in all 35 volunteers--findings consistent with those of du Boulay. However, contrary to du Boulay's observations, asymmetric loss of signal, consistent with pulsatile CSF flow, was demonstrated at the level of the foramen of Monro in 15 of the 35 volunteers. Based on the pattern of flow void at the level of the foramen of Monro and on the expansion of the third ventricle during systole, we propose a theory of synchronous CSF flow at the foramen of Monro and aqueduct, which unifies our MR findings with du Boulay's cineventriculographic observations.

Functional magnetic resonance imaging. Application to degenerative brain disease and hydrocephalus.

Magnetic resonance (MR) imaging, in addition to its excellent depiction of neuroanatomy, is being developed as a major technique for functional imaging of cerebrospinal fluid motion and for measurement of velocity, strain, and diffusional processes within the brain parenchyma. Functional MR imaging studies are revealing basic physiology of blood flow interactions with CSF motion and dynamic processes of brain parenchyma. Normal pressure hydrocephalus and degenerative brain disease are current focuses of functional MR imaging studies.

VET imaging: magnetic resonance imaging with variable encoding time.

A methodology of MRI data acquisition is introduced which involves lengthening the duration of signal readout period with complementary shortening of the phase encode pulses, and vice versa, in sequences where gradient encoding time is limited by RF pulse spacing. This variable encoding time (VET) methodology can be used to increase spatial resolution or reduce data acquisition time in 2D and 3D FT MRI based on CPMG and gradient echo sequences. Advantages of higher image SNR and different spatial frequency distributions in k-space are discussed and evaluated in preliminary experiments.

GRASE (gradient- and spin-echo) MR imaging: a new fast clinical imaging technique.

A novel technique of magnetic resonance (MR) imaging, which combines gradient-echo and spin-echo (GRASE) technique, accomplishes T2-weighted multisection imaging in drastically reduced imaging time, currently 24 times faster than spin-echo imaging. The GRASE technique maintains contrast mechanisms, high spatial resolution, and image quality of spin-echo imaging and is compatible with clinical whole-body MR systems without modification of gradient hardware. Image acquisition time is 18 seconds for 11 multisection body images (2,000/80 [repetition time msec/echo time msec]) and 36 seconds for 22 brain images (4,000/104). With a combination of multiple Hahn spin echoes and short gradient-echo trains, the GRASE technique overcomes several potential problems of echo-planar imaging, including large chemical shift, image distortions, and signal loss from field inhomogeneity. Advantages of GRASE over the RARE (rapid acquisition with relaxation enhancement) technique include faster acquisition times and lower deposition of radio-frequency power in the body. Breath holding during 18-second GRASE imaging of the upper abdomen eliminates respiratory-motion artifacts in T2-weighted images. A major improvement in T2-weighted abdominal imaging is suggested.

GRASE (Gradient- and spin-echo) imaging: a novel fast MRI technique.

A fast multi-section MR imaging technique is described. Gradient- and spin-echo (GRASE) imaging utilizes the speed advantages of gradient refocusing while overcoming the image artifacts arising from static field inhomogeneity and chemical shift. Image contrast is determined by the T2 contrast in the Hahn spin echoes. A novel k-space trajectory temporally modulates signals and demodulates artifacts.

A methodology for co-registering abdominal MR images over multiple breath-holds.

Previous studies have demonstrated that the SNR of abdominal MR images can be increased by averaging images obtained in different breath-hold acquisitions. In this note, the authors present a simple new methodology for ensuring that images acquired in multiple breath-hold periods are accurately co-registered. Within each breath-hold, a quick coronal scout scan is followed by a longer axial scan. The scout is used to position the axial slices in a fixed position relative to the organ under examination. This MR technique can, in principle, be automated so as to add less than 1 s to the imaging time of the axial scan. The method can be used to increase SNR by signal averaging or to co-register images acquired during, for example, uptake of contrast agents. SNR improvement with negligible blurring is demonstrated in liver images acquired by this method from healthy volunteers.

Phase errors in multi-shot echo planar imaging.

Field inhomogeneity related phase errors in multi-shot echo planar imaging (EPI) are directly visualized and analyzed in the spatial frequency domain data or 'k-space'. The echo time shift (ETS) technique incrementally moves the position of the echo train and improves the phase error function by redistributing phase discontinuities away from the center of k-space.

Human brain motion and cerebrospinal fluid circulation demonstrated with MR velocity imaging.

Present theory holds that pulsatile pressure of cerebrospinal fluid (CSF) is driven by the force of expansion of the choroid plexus. Alternate theories postulating that a possible movement of the brain is involved in pumping CSF have not, to the authors' knowledge, been substantiated heretofore. In this study, in vivo, quantitative magnetic resonance (MR) imaging methods were developed to show reproducible magnitudes and directions of CSF flow. Measurements were obtained with a new MR velocity imaging technique at high resolution (0.4 mm/sec), requiring 64 cardiac cycles per image. Twenty-five healthy volunteers and five patients were studied. Observations of pulsatile brain motion, ejection of CSF out of the cerebral ventricles, and simultaneous reversal of CSF flow direction in the basal cisterns toward the spinal canal, taken together, suggest that a vascular-driven movement of the entire brain may be directly pumping the CSF circulation. The authors describe what they believe to be the first observations and measurements of human brain motion, which occurs in extensive internal regions (particularly the diencephalon and brain stem) and is synchronous with cardiac systole.

Tissue perfusion in humans studied by Fourier velocity distribution, line scan, and echo-planar imaging.

In tissue perfusion studies, FT velocity distribution imaging (VDI) intrinsically distinguishes signals from moving blood and volume-averaged tissue. Results in human thyroid gland, in vivo, using VDI line scan technique demonstrated separation of moving blood signal from glandular tissue, while VDI inner-volume echo-planar imaging of brain showed only CSF velocity above the image noise level. New alternating polarity gradient sequences which permit separation of diffusion and slow velocity are discussed. A novel method of 3D FT imaging (two spatial and one velocity dimension) combining inner-volume imaging and echo-planar imaging with velocity resolution of 0.15 mm/s per pixel is demonstrated. A novel graphical method of calculation and display of diffusion dependence in pulsed gradient sequences is presented.

Single-shot GRASE imaging without fast gradients.

Based on the CPMG sequence, gradient- and spin-echo (GRASE) echo train length is limited by T2 decay rather than the T2* decay and phase error in echo-planar techniques, permitting a longer image acquisition period. An ultrafast GRASE sequence, utilizing a single excitation, generates a 128 x 56 true T2-weighted image in 200 ms on an unmodified commercial scanner without fast gradient switching, extreme field homogeneity, or fat signal suppression.

Multiple breath-hold averaging (MBA) method for increased SNR in abdominal MRI.

Breath-holding during MR imaging eliminates respiratory motion artifacts but places a major time constraint on data acquisition. This constraint limits image signal-to-noise ratio and hence spatial resolution. A new method, multiple breath-hold averaging, is presented that overcomes these time limitations. Several images are acquired in sequential breath-hold periods, separated by periods of normal breathing, and averaged. This averaged image shows the expected increase in SNR with surprisingly little blurring due to misregistration. SNR improvements can be traded for increased spatial resolution. The MBA methodology can also be applied to 3D data acquisitions, dynamic contrast acquisitions, and image subtractions.

High resolution GRASE MRI of the brain and spine: 512 and 1024 matrix imaging.

OBJECTIVE: Our goal was to implement 512 and 1024 matrix MRI of the brain in clinically acceptable imaging times. MATERIALS AND METHODS: With use of the GRASE (gradient-SE) imaging technique, 3 signals are refocused from each of 16 SEs, giving a total of 48 echoes per echo train, for a speed advantage of 1/48 over conventional SE imaging. Images in 1024 matrices are acquired in 2D Fourier transform (FT) multisectional MRI. In related experiments, multislab 3D FT GRASE imaging is performed using 16 partitions/slab, providing thinner 1 to 2 mm sections. In all imaging experiments, higher spatial resolution is obtained with stronger and faster gradients, 24 mT/m in 625 microseconds rise time, on a standard commercial MRI system. RESULTS: The 2D FT 1024 matrix images of the head were acquired in 4:20 min, with 20 sections and TR/TE 7040/115 ms in a rectangular FOV to obtain .28 x .27 mm2 spatial resolution. Small anatomic structures, including cochlea of inner ear, cranial nerves, and vascular detail, are readily demonstrated. The 512 matrix images were obtained in 4:40 min, with 16 sections and TR/TE 3,500/104 ms in a 24 cm FOV. The 3D FT technique substantially increased slice coverage as well as image signal-to-noise ratio. CONCLUSION: The results show that 1024 matrix MRI is technically feasible in clinically acceptable imaging time and offers advantage for high resolution imaging. Optimization of 1024 matrix and 3D FT GRASE imaging should improve the delineation of anatomic regions of interest.

Dual contrast GRASE (gradient-spin echo) imaging using mixed bandwidth.

Equal time spacing of RF pulses in the CPMG sequence imposes a constraint of equal signal read periods in spin-echo train imaging. GRASE imaging differs by using multiple read gradients in each pi-pi time interval, which are not constrained to be equal in number or duration. This additional degree of freedom is developed in dual contrast imaging. Closely spaced read periods are used for the PDW image to reduce T2 decay effects, while fewer low-bandwidth read periods in each of several pi-pi intervals are used to raise the signal-to-noise ratio and avoid signal averaging in the T2-weighted image.

Increased flexibility in GRASE imaging by k space-banded phase encoding.

GRASE (GRadient and spin Echo) is an echo train imaging technique that combines gradient and RF refocusing. Although overall signal decay is with T2 and field inhomogeneity phase errors do not accumulate, the small residual phase errors are periodic with echo number. The echo order described previously eliminates the phase error periodicity in k space but instead creates periodicity in the T2 modulation function that can also cause artifacts. In addition, with this order, the effective TE must be half the echo train time, and asymmetric Fourier sampling is difficult to implement. A new method is described that greatly reduces artifacts due to T2 decay, permits greater control of T2 contrast, and lends itself to asymmetric Fourier sampling. Different time segments of the echo train are encoded with different bands of spatial frequency in k space (hence "k banding"). Both computer simulations and experimental results demonstrate improvements in GRASE images acquired by this method.

GRASE improves spatial resolution in single shot imaging.

In single shot echo train imaging all the data required for a two dimensional image is acquired from a series of echoes generated following a single RF excitation pulse. Spatial resolution is limited because all samples must be acquired before the signal decays. In this paper we show theoretically that more echoes and hence better spatial resolution can be obtained with single shot GRASE imaging than with either echo planar imaging or single shot RARE imaging. This conclusion holds for both conventional imaging hardware and specialized gradient hardware designed for EPI. High quality single shot GRASE images support the theoretical conclusions.

Single-shot GRASE imaging with short effective TEs.

A new phase-encoding scheme for gradient- and spin-echo (GRASE) imaging giving a short effective TE is described. Unlike previous orders, phase encoding is centric rather than sequential. The sequence is a development of k-banded GRASE that uses different time segments of the echo train to encode different bands of k space. This phase-encoding order has been implemented in single-shot sequences on an imager with high performance gradients. Approximately 144 phase-encoding lines can be acquired in an echo train time of 390 ms. With centric phase encoding, the effective TE is 8 ms, compared with 75 ms for sequential encoding, and signal-to-noise ratios (SNRs) in brain tissue are 50 to 70% higher. The sequence can be employed in, for example, diffusion and velocity imaging.

A comparison of phase encoding ordering schemes in T2-weighted GRASE imaging.

Gradient and spin echo (GRASE) imaging is an echo train imaging sequence that combines gradient and RF refocusing. This combination introduces phase modulations into the echo train. If the phase encoding order is linear with echo time, these modulations cause severe ghosting artifacts. Changing the order of phase encoding can greatly reduce these artifacts. Several phase encoding orders for T2-weighted sequences are compared in this paper, linear, partially randomized, standard GRASE ordering, and k-banded (kb) GRASE ordering. Different possible implementations of GRASE and kbGRASE are also considered. Computer simulation is used to compare resolution and artifact levels. Phantom and volunteer images are presented. The linear order is most sensitive to ghosting artifacts associated with chemical shift, susceptibility differences and static field inhomogeneities. The standard GRASE order is least sensitive to these but most vulnerable to artifacts associated with short T2 signals, kb-GRASE is a good intermediate between linear and standard GRASE and generally shows the lowest artifact levels. The partially randomized order gives the most diffuse artifacts. Computer simulations show that spatial resolution and contrast with all phase encoding orders are similar.