Noise performance of a precision pulsed electromagnet power supply for magnetic resonance imaging

Prepolarized magnetic resonance imaging (PMRI) uses two pulsed electromagnets to achieve high-field image quality with the benefits of low-field data acquisition. The principal challenge with all resistive MRI systems is the implementation of a highly precise magnet current supply. The noise current through the magnet is fundamentally limited by the current transducer used to provide feedback and the voltage reference used to generate the demand signal. Field instability in the main field magnet can both corrupt the received data and degrade the robustness of Carr¿Purcell¿Meiboom¿Gill (CPMG) echo trains, which are paramount to efficient imaging in PMRI. In this work, we present the magnet control system that achieved sufficient field stability for PMRI at $0.5/0.13$ T, identify the dominant sources of noise in the control system, examine the imaging artifacts that can occur if the field stability is insufficient, and identify how the design can be improved for better field stability, should it be required for future implementations of PMRI.

Rapid evaluation of left ventricular volume and mass without breath-holding using real-time interactive cardiac magnetic resonance imaging system

OBJECTIVES

The purpose of this study was to validate cardiac measurements derived from real-time cardiac magnetic resonance imaging (MRI) as compared with well-validated conventional cine MRI.

BACKGROUND

Although cardiac MRI provides accurate assessment of left ventricular (LV) volume and mass, most techniques have been relatively slow and required electrocardiogram (ECG) gating over many heart beats. A newly developed real-time MRI system allows continuous real-time dynamic acquisition and display without cardiac gating or breath-holding.

METHODS

Fourteen healthy volunteers and nine patients with heart failure underwent real-time and cine MRI in the standard short-axis orientation with a 1.5T MRI scanner. Nonbreath-holding cine MRI was performed with ECG gating and respiratory compensation. Left ventricular end-diastolic volume (LVEDV), left ventricular endsystolic volume (LVESV), ejection fraction (EF) and LV mass calculated from the images obtained by real-time MRI were compared to those obtained by cine MRI.

RESULTS

The total study time including localization for real-time MRI was significantly shorter than cine MRI (8.6 +/- 2.3 vs. 24.7 +/- 3.5 min, p < 0.001). Both imaging techniques yielded good quality images allowing cardiac measurements. The measurements of LVEDV, LVESV, EF and LV mass obtained with real-time MRI showed close correlation with those obtained with cine MRI (LVEDV: r = 0.985, p < 0.001; LVESV: r = 0.994, p < 0.001; EF: r = 0.975, p < 0.001; LV mass: r = 0.977, p < 0.001).

CONCLUSIONS

Real-time MRI provides accurate measurements of LV volume and mass in a time-efficient manner with respect to image acquisition.

The real-time interactive 3-D-DVA for robust coronary MRA

A graphical user interface (GUI) has been developed which enables interactive feedback and control to the real-time diminishing variance algorithm (DVA). This interactivity allows the user to set scan parameters, view scan statistics, and view image updates during the course of the scan. In addition, the DVA has been extended to simultaneously reduce motion artifacts in three dimensions using three orthogonal navigators. Preliminary in vivo studies indicate that these improvements to the standard DVA allow for significantly improved consistency and robustness in eliminating respiratory motion artifacts from MR images, particularly when imaging the coronary arteries.

Nonsubtractive spiral phase contrast velocity imaging

Phase contrast velocity imaging is a standard method for accurate in vivo flow measurement. One drawback, however, is that it lengthens the scan time (or reduces the achievable temporal resolution) because one has to acquire two or more images with different flow sensitivities and subtract their phases to produce the final velocity image. Without this step, non-flow-related phase variations will give rise to an erroneous, spatially varying background velocity. In this paper, we introduce a novel phase contrast velocity imaging technique that requires the acquisition of only a single image. The idea is to estimate the background phase variation from the flow-encoded image itself and then have it removed, leaving only the flow-related phase to generate a corrected flow image. This technique is sensitive to flow in one direction and requires 50% less scan time than conventional phase contrast velocity imaging. Phantom and in vivo results were obtained and compared with those of the conventional method, demonstrating the new method's effectiveness in measuring flow in various vessels of the body. Magn Reson Med 42:704-713, 1999.

Resistive homogeneous MRI magnet design by matrix subset selection

A new technique for designing resistive homogeneous multicoil magnets for magnetic resonance imaging (MRI) is presented. A linearly independent subset of coils is chosen from a user-defined feasible set using an efficient numerical algorithm. The coil currents are calculated using a linear least squares algorithm to minimize the deviation of the actual magnetic field from the target field. The solutions are converted to practical coils by rounding the currents to integer ratios, selecting the wire gauge, and optimizing the coil cross-sections. To illustrate the technique, a new design of a short, homogeneous MRI magnet suitable for low-field human torso imaging is presented. Magnets that satisfy other constraints on access and field uniformity can also be designed. Compared with conventional techniques that employ harmonic expansions, this technique is flexible, simple to implement, and numerically efficient.

New real-time interactive cardiac magnetic resonance imaging system complements echocardiography

OBJECTIVES

We conducted an initial clinical trial of a newly developed cardiac magnetic resonance imaging (CMRI) system. We evaluated left ventricular (LV) function in 85 patients to compare the clinical utility of the CMRI system with echocardiography, the current noninvasive gold standard.

BACKGROUND

Conventional CMRI systems require cardiac-gating and respiratory compensation to synthesize a single image from data acquired over multiple cardiac cycles. In contrast, the new CMRI system allows continuous real-time dynamic acquisition and display of any scan plane at 16 images/s without the need for cardiac gating or breath-holding.

METHODS

A conventional 1.5T Signa MRI Scanner (GE, Milwaukee, Wisconsin) was modified by the addition of an interactive workstation and a bus adapter. The new CMRI system underwent clinical trial by testing its ability to evaluate global and regional LV function. The first group (A) consisted of 31 patients with acceptable echocardiography image quality. The second group (B) consisted of 31 patients with suboptimal echocardiography image quality. The third group (C) consisted of 29 patients with severe lung disease or congenital cardiac malformation who frequently have suboptimal echo study. Two independent observers scored wall motion and image quality using the standard 16-segment model and rank-order analysis.

RESULTS

CMRI evaluation was complete in less than 15 min. In group A, no significant difference was found between ECHO and CMRI studies (p = NS). In group B, adequate visualization of wall segments was obtained 38% of the time using ECHO and 97% of the time using CMRI (p < 0.0001). When grouped into coronary segments, adequate visualization of at least one segment occurred in 18 of 30 patients (60%) with ECHO and in all 30 patients (100%) with CMRI (p < 0.0001). In group C, adequate visualization of the wall segments was obtained in 58% (CI 0.53-0.62) of the time using echocardiography and 99.7% (CI 0.99-1.0) of the time using CMRI (p < 0.0001).

CONCLUSIONS

The new CMRI system provides clinically reliable evaluation of LV function and complements suboptimal echocardiography. In comparison with the conventional CMRI, the new CMRI system significantly reduces scan time, patient discomfort and associated cost.

Magnetic resonance imaging of knee cartilage repair

Cartilage injury resulting in osteoarthritis is a frequent cause of disability in young people. Osteoarthritis, based on either cartilage injury or degeneration, is a leading cause of disability in the United States. Over the last several decades, much progress has been made in understanding cartilage injury and repair. Magnetic resonance (MR) imaging, with its unique ability to noninvasively image and characterize soft tissue, has shown promise in assessment of cartilage integrity. In addition to standard MR imaging methods, MR imaging contrast mechanisms under development may reveal detailed information regarding the physiology and morphology of cartilage. MR imaging will play a crucial role in assessing the success or failure of therapies for cartilage injury and degeneration.

Feasibility study of lactate imaging of head and neck tumors

A proton spectroscopic imaging sequence was used to investigate the feasibility of lactate imaging in head and neck tumors. The sequence employs a two-shot lactate editing method with inversion recovery for additional lipid suppression, and a restricted field of view to suppress motion artifacts. Variations in acquisition parameters and two different receive coils were investigated on twelve patients. Elevated lactate was detected in three patients, no lactate was observed in seven patients, and two studies were inconclusive because of severe motion or inhomogeneity artifacts. Best results were obtained with an anterior/posterior neck coil at a 288 ms echo time (TE).

Volumetric spectroscopic imaging with spiral-based k-space trajectories

Spiral-based k-space trajectories were applied in a spectroscopic imaging sequence with time-varying readout gradients to collect volumetric chemical shift information. In addition to spectroscopic imaging of low signal-to-noise ratio (SNR) brain metabolites, the spiral trajectories were used to rapidly collect reference signals from the high SNR water signal to automatically phase the spectra and to aid the reconstruction of metabolite maps. Spectral-spatial pulses were used for excitation and water suppression. The pulses were designed to achieve stable phase profiles in the presence of up to 20% variation in the radiofrequency field. A gridding algorithm was used to resample the data onto a rectilinear grid before fast Fourier transforms. This method was demonstrated by in vivo imaging of brain metabolites at 1.5 T with 10 slices of 18 x 18 pixels each. Nominal voxel size was 1.1 cc, spectral bandwidth was 400 Hz, scan time was 18 min for the metabolite scan and 3 min for the reference scan.

Real-time interactive MRI on a conventional scanner

A real-time interactive MRI system capable of localizing coronary arteries and imaging arrhythmic hearts in real-time is described. Non-2DFT acquisition strategies such as spiral-interleaf, spiral-ring, and circular echo-planar imaging provide short scan times on a conventional scanner. Real-time gridding reconstruction at 8-20 images/s is achieved by distributing the reconstruction on general-purpose UNIX workstations. An X-windows application provides interactive control. A six-interleaf spiral sequence is used for cardiac imaging and can acquire six images/s. A sliding window reconstruction achieves display rates of 16-20 images/s. This allows cardiac images to be acquired in real-time, with minimal motion and flow artifacts, and without breath holding or cardiac gating. Abdominal images are acquired at over 2.5 images/s with spiral-ring or circular echo-planar sequences. Reconstruction rates are 8-10 images/s. Rapid localization in the abdomen is demonstrated with the spiral-ring acquisition, whereas peristaltic motion in the small bowel is well visualized using the circular echo-planar sequence.

Lesion contrast enhancement in medical ultrasound imaging

Methods for improving the contrast-to-noise ratio (CNR) of low-contrast lesions in medical ultrasound imaging are described. Differences in the frequency spectra and amplitude distributions of the lesion and its surroundings can be used to increase the CNR of the lesion relative to the background. Automated graylevel mapping is used in combination with a contrast-weighted form of frequency-diversity speckle reduction. In clinical studies, the techniques have yielded mean CNR improvements of 3.2 dB above ordinary frequency-diversity imaging and 5.6 dB over sharper conventional images, with no post-processing graylevel mapping.

Improved automatic off-resonance correction without a field map in spiral imaging

Non-2DFT k-space readout strategies are useful in fast imaging but prone to blurring when reconstructed off resonance. Field inhomogeneities or susceptibility variations, coupled with a long readout time, are the major sources of this artifact. Correction methods based on a priori off-resonance information such as an acquired field map have been proposed in the literature. An alternative approach estimates the spatially varying off-resonance frequency from the data itself before applying a correction. In this latter approach there is a trade-off between the extent of correction and the chance of increased artifact due to estimation error. This paper introduces an improved algorithm for field map estimation which is both faster and more robust than the existing method. It uses a multi-stage estimation of the field map, starting from a coarse estimate both in frequency and space and proceeds towards higher resolution. The new algorithm is applied to phantom and in vivo images acquired with radial and spiral sequences to give sharper images.

Multifrequency interpolation for fast off-resonance correction

Field inhomogeneities or susceptibility variations produce blurring in images acquired using non-2DFT k-space readout trajectories. This problem is more pronounced for sequences with long readout times such as spiral imaging. Theoretical and practical correction methods based on an acquired field map have been reported in the past. This paper introduces a new correction method based on the existing concept of frequency segmented correction but which is faster and theoretically more accurate. It consists of reconstructing the data at several frequencies to form a set of base images that are then added together with spatially varying linear coefficients derived from the field map. The new algorithm is applied to phantom and in vivo images acquired with projection reconstruction and spiral sequences, yielding sharply focused images.

Three-dimensional spectral-spatial excitation

A method of three-dimensional spectral-spatial excitation is presented which is selective simultaneously in two spatial dimensions and in the spectral or chemical shift dimension. This method can be used to create spectral passbands whose center frequency varies as a function of spatial location within an imaging plane. This variation of passband center frequency may be specified by an acquired main field (B0) map; the resulting excitations compensate for inhomogeneity of the B0 field. In vivo images are presented in which three-dimensional spectral-spatial excitation allows selective water-only imaging in the presence of large B0 inhomogeneity where conventional spectrally selective imaging falls. Phantom studies give a detailed profile of the performance of three-dimensional spectral-spatial pulses suitable for water-only or fat-saturation imaging. These pulses may also be useful for fat and water suppression in spectroscopic imaging. Performance constraints imposed by limited gradient slew rates are analyzed and quantified.

A readout magnet for prepolarized MRI

Conventional MRI systems rely on large magnets to generate a field that is both strong and extremely uniform. This field is usually produced by a heavy permanent magnet or a cryogenically cooled superconductor. An alternative approach, called prepolarized MRI (PMRI), employs two separate fields produced by two different magnets. A strong and inhomogeneous magnetic field is used to polarize the sample. After polarization, a weak magnetic field is used for readout. These fields can be produced by two separate resistive electromagnets that cost significantly less than a single permanent or superconducting magnet. At Stanford, the authors are constructing a PMRI prototype scanner suitable for imaging human extremities roughly 20 cm in diameter. With this system the authors hope to demonstrate comparable image quality to MRI with reduced system cost. The authors' initial work on low-frequency reception indicates that it will be possible to obtain comparable image signal-to-noise ratio to an MRI scanner operating at the same polarizing field strength. To reduce the capital cost of the system, the authors use resistive electromagnets. Here the authors discuss the full development of the readout magnet including important design considerations, shimming, and field plots. These encouraging results are an important step toward evaluating the cost effectiveness of PMRI.

Noise in MRI

This study analyzes the signal-to-noise ratio (SNR) in magnetic resonance imaging. The factors that determine the SNR are derived starting from basic principles. The SNR, for a given object, is shown to be proportional to the voxel volume and the square root of the acquisition time. The noise generated by the body is derived using a cylindrical model and is shown to be proportional to the square of the radius and the square root of the length.

Inhomogeneity correction using an estimated linear field map

A fast and robust method for correcting magnetic resonance image distortion due to field inhomogeneity is proposed and applied to spiral k-space scanning. The method consists of acquiring a local field map, finding the best fit to a linear map, and using it to deblur the image distortions due to local frequency variations. The linear field map is determined using a maximum likelihood estimator with weights proportional to the pixel intensity. The method requires little additional computation and is robust in low signal regions and near abrupt field changes. Additionally, it can be used in combination with other deblurring methods. The application of this method is illustrated in conjunction with a multislice, T2-weighted, breath-held spiral scan of the liver.

MRS imaging using anatomically based k-space sampling and extrapolation

A comprehensive strategy for the acquisition, reconstruction, and postprocessing of MR spectroscopic images is presented. The reconstruction algorithm is the most critical component of this strategy. It is assumes that the desired image is spatially bounded, meaning that the desired image contains an object that is surrounded by a background of zeros. The reconstruction algorithm relies on prior knowledge of the background zeros for k-space extrapolation. This algorithm is a good candidate for proton MR spectroscopic image reconstruction because these images are often spatially bounded and prior knowledge of the zeros is easily obtained from a rapidly acquired high resolution conventional MRI. Although the reconstruction algorithm can be used with the standard 3DFT k-space distribution, a distribution that relies on anatomical features that are likely to occur in the spectroscopic image can produce better results. Prior knowledge of these anatomical features is also obtained from a conventional MRI. Finally, the postprocessing component of this strategy is valuable for reducing subcutaneous lipid contamination. Overall, the comprehensive approach presented here produces images that are better resolved than standard approaches without increasing acquisition time or reducing SNR. Examples using NAA data are provided.

MR spectroscopic imaging of collagen: tendons and knee menisci

Water molecules associated with collagen have short transverse (T2) relaxation times. Projection-reconstruction techniques are able to achieve an echo time (TE) much shorter than conventional techniques, allowing imaging of tissues with T2 < 5 ms. Using these techniques, a conventional 1.5-T MRI human imaging system can directly image collagen-associated water from knee menisci and tendons in normal volunteers and patients. Long-T2 suppression improves the contrast between these structures and the surrounding tissue with long-T2 relaxation times. Spectroscopic imaging provides improved lipid/water registration and information about chemical composition and relaxation times. Direct imaging of tendons and menisci may provide more information about these structures and provide a new way to assess both injury and repair.

The diminishing variance algorithm for real-time reduction of motion artifacts in MRI

A technique has been developed whereby motion can be detected in real time during the acquisition of data. This enables the implementation of several algorithms to reduce or eliminate motion effects from an image as it is being acquired. One such algorithm previously described is the acceptance/rejection method. This paper deals with another real-time algorithm called the diminishing variance algorithm (DVA). With this method, a complete set of preliminary data is acquired along with information about the relative motion position of each frame of data. After all the preliminary data are acquired, the position information is used to determine which data frames are most corrupted by motion. Frames of data are then reacquired, starting with the most corrupted one. The position information is continually updated in an iterative process; therefore, each subsequent reacquisition is always done on the worst frame of data. The algorithm has been implemented on several different types of sequences. Preliminary in vivo studies indicate that motion artifacts are dramatically reduced.

Coronary angiography with magnetization-prepared T2 contrast

A magnetization-prepared, T2-weighted sequence (T2 Prep) is used to suppress muscle and venous structures. When combined with lipid suppression, this technique improves the visualization of the coronary arteries. T2 Prep was designed to be rebust in the presence of flow as well as B0 and B1 inhomogeneities and may be combined with virtually any imaging technique. Here, it is implemented with both a single-slice spiral acquisition and a multi-slice spiral method that acquires up to 15 slices in a single breath-holding interval.

Three-dimensional spectroscopic imaging with time-varying gradients

A spectroscopic imaging sequence with a time-varying readout gradient in the slice selection direction is used to image multiple contiguous slices. For a given voxel size, the imaging time and signal-to-noise ratio of the three-dimensional spectroscopic sequence are the same as for a single slice acquisition without the oscillating readout gradient. The data reconstruction employs a gridding algorithm in two dimensions to interpolate the nonuniformly sampled data onto a Cartesian grid, and a fast Fourier transform in four dimensions: three spatial dimensions and the spectral dimension. The method is demonstrated by in vivo imaging of NAA in human brain at 1.5 T with 10 slices of 16 x 16 pixels spectroscopic images acquired in a total scan time of 17 min.

Spectral extrapolation of spatially bounded images [MRI application]

A spectral extrapolation algorithm for spatially bounded images is presented. An image is said to be spatially bounded when it is confined to a closed region and is surrounded by a background of zeros. With prior knowledge of the spatial domain zeros, the extrapolation algorithm extends the image's spectrum beyond a known interval of low-frequency components. The result, which is referred to as the finite support solution, has space variant resolution; features near the edge of the support region are better resolved than those in the center. The resolution of the finite support solution is discussed as a function of the number of known spatial zeros and known spectral components. A regularized version of the finite support solution is included for handling the case where the known spectral components are noisy. For both the noiseless and noisy cases, the resolution of the finite support solution is measured in terms of its impulse response characteristics, and compared to the resolution of the zerofilled and Nyquist solutions. The finite support solution is superior to the zerofilled solution for both the noisy and noiseless data cases. When compared to the Nyquist solution, the finite support solution may be preferred in the noisy data case. Examples using medical image data are provided.

Real-time motion detection in spiral MRI using navigators

A technique has been developed whereby motion can be detected in real time during the acquisition of data. This enables the implementation of an algorithm to accept or reject and reacquire data during a scan. Frames of data with motion are rejected and reacquired on the fly so that by the end of the scan, a complete motion-free data set has been acquired. The algorithm has been implemented on several different types of sequences. Preliminary in vivo studies indicate that motion artifacts are dramatically reduced.

Short TE phosphorus spectroscopy using a spin-echo pulse

In vivo phosphorus spectroscopy requires very short acquisition delays in order to capture the signal from components with short transverse relaxation times (T2). The echo time typical of standard slice selective spin-echo pulses are too long for this application, so hard pulse, free induction decay (FID) acquisitions have frequently been used instead. With FID, however, there is an interval between the time of coherence and data acquisition (acquisition delay), with resulting baseline distortions. In this paper we describe the design of a new short TE, slice-selective, composite spin-echo pulse with echo times as short as 2.5 ms. With a long TR, fully relaxed, multislice spectra can be collected. This technique will be useful for assessing in vivo, changes in brain phospholipid activity associated with psychiatric and neurological diseases.

Magnetic resonance spectroscopic imaging of ethanol in the human brain: a feasibility study

The in vivo distribution of ethanol in normal human brain following the consumption of a moderate amount of alcohol was measured using magnetic resonance spectroscopic imaging. Three subjects were studied, and the spatial distribution of brain ethanol, 60-min postingestion and measured at a spatial resolution of 1.5 cm, was found to be highly nonuniform with the relative ethanol signal in cerebral spinal fluid, gray matter, and white matter determined to be 1.00, 0.72, and 0.37, respectively. These spectroscopic imaging results indicate that whereas in vivo magnetic resonance studies of ethanol are feasible, quantitative studies of alcohol need to account carefully for the various tissue types within the observed volume.

Novel approaches to low-cost MRI

This paper presents a combination of speculative approaches, some related to earlier work and some apparently novel, which show great promise in providing a new class of MRI machines that would be considerably less expensive. This class would have advantages and disadvantages as compared to existing MRI, over and above that of low cost. The disadvantages include the apparent inability to perform classic spectroscopy, and limited flexibility in the area of selective excitation. The advantages include a fundamental immunity to inhomogeneity and susceptibility problems, the ability to create a wide class of machines that are designed for specific anatomy-related applications, the ability to design open machines for physician access, and improved capability for high speed imaging. Generic to all of the methods presented are a pulsed polarizing field and an oscillatory read-out bias field. The pulsed field initially polarizes the magnetic moments. Since it is not on during the readout operation it has negligible homogeneity requirements since changes in the field amplitude will merely shade the image intensity. During readout a relatively low bias field is used. To enable the use of a relatively inhomogeneous bias field, an oscillatory field is used that has a zero average value. This prevents any long-term buildup of phase errors due to a frequency error associated with inhomogeneity. Thus the average bias frequency will be determined solely by the frequency rather than the amplitude of the bias field. Three methods are described, all including the above features. The first two involve imaging in the laboratory frame, while the third involves imaging in the rotating frame. The second approach requires no RF excitation and the third approach uses RF bias and gradient signals. Some approaches to slice selection are described.

Incorporating lactate/lipid discrimination into a spectroscopic imaging sequence

A spectroscopic imaging sequence incorporating a two-shot lactate editing method was used in two human brain studies to image lactate and NAA. The subtractive editing method allows separate images of lactate, NAA, and lipids to be collected during a single study with no SNR penalty. The sequence uses a spectral-spatial excitation for slice selection and water suppression, and employs inversion recovery and an echo time of 136 ms for additional lipid suppression.

Echo-planar spin-echo and inversion pulses

The echo-planar k-space trajectory can be used as the basis for any two-dimensional selective pulse. The main application is spectral-spatial pulses, which must be based on the echo-planar trajectory. In this paper we show how echo-planar spin-echo (EPSE) pulses may be designed.

Characterization of atherosclerosis with a 1.5-T imaging system

It is shown that a conventional 1.5-T magnetic resonance (MR) imaging system can help characterize some of the key components of atherosclerotic plaque ex vivo. Fresh human aorta with atheromata was suspended in solutions of agarose and manganese chloride and heated to body temperature. The specimens were imaged with modified Dixon and projection-reconstruction imaging sequences. The specimens were then examined histologically to obtain direct correlation between images, spectra, and histologic characteristics. The results show that vessel wall and plaque components can be identified by means of their MR characteristics and correlated with their histologic appearance. The authors were able to identify normal vessel wall components, such as adventitial lipids and smooth muscle. They were also able to identify and localize plaque components such as fibrous tissue, calcification, lipids, and possible areas of hemorrhage and hemosiderin deposition.

Prospects for ultrasonic spectroscopy and spectral imaging of abdominal tissues

A system for the digitization and frequency spectral analysis of radiofrequency data for ultrasonic waveforms backscattered from abdominal tissues is described. Studies of phantoms meant to simulate abdominal tissues of differing scattering characteristics indicated that frequency spectral differences due to differences in the frequency dependence of backscattering were seen with 5 MHz probes, but not with a 3.5 MHz probe. Studies of a phantom with a simulated lesion of altered scattering characteristics indicated potential for improved lesion detection and characterization, using custom circuitry developed for variable bandwidth filtering of received ultrasonic beams. The techniques discussed have potential for improved diagnosis of diffuse and focal abdominal abnormalities over that obtained with conventional ultrasonic imaging.

A three-dimensional spin-echo or inversion pulse

In theory, multidimensional pulses can be designed to be selective in any number of dimensions. In practice, available gradient power has enforced a limit to two dimensions. We show here that three-dimensional pi pulses are feasible on commercial imaging machines provided that the range of off-resonance frequencies are limited.

Fast spiral coronary artery imaging

A flow-independent method for imaging the coronary arteries within a breath-hold on a standard whole-body MR imager was developed. The technique is based on interleaved spiral k-space scanning and forms a cardiac-gated image in 20 heartbeats. The spiral readouts have good flow properties and generate minimal flow artifacts. The oblique slices are positioned so that the arteries are in the plane and so that the chamber blood does not obscure the arteries. Fat suppression by a spectral-spatial pulse improves the visualization of the arteries.

Pulsed saturation transfer contrast

In vivo 1H conventional NMR image contrast generation usually relies on the macroscopic T1 and T2 relaxation parameters of the tissues of interest. Recently cross-relaxation related image contrast has been reported by Wolff and Balaban in animal models. Due primarily to the broad lineshape of the intended saturation spin pool and the use of off-resonance irradiation, high specific absorption rate and an auxiliary RF amplifier have been necessary to produce these images. The relatively long spin-lattice relaxation property of this spin pool, however, suggests the use of pulse methods to achieve saturation. In this paper, we show that short-T2 spin pools can be selectively saturated with short intense RF pulses. Cross-relaxation time constants can be measured using the technique of saturation recovery. In vivo magnetization-transfer-weighted images can be produced using pulses on commercial whole-body imagers without additional hardware.

Deblurring for non-2D Fourier transform magnetic resonance imaging

For several non-2D Fourier transform imaging methods, off-resonant reconstruction does not just cause geometric distortion, but changes the shape of the point spread function and causes blurring. This effect is well known for projection reconstruction and spiral k-space scanning sequences. We introduce here a method that automatically removes blur introduced by magnetic field inhomogeneity and susceptibility without using a resonant frequency map, making these imaging methods more useful. In this method, the raw data are modulated to several different frequencies and reconstructed to create a series of base images. Determination of degree of blur is done by calculating a focusing measure for each point in each base image and a composite image is then constructed using only the unblurred regions from each base image. This method has been successfully applied to phantom and in vivo images using projection-reconstruction and spiral-scan sequences.

MR imaging of lung parenchyma: a solution to susceptibility

The authors have developed a pulse sequence for imaging lung parenchyma with projection reconstruction magnetic resonance (MR) imaging that reduces the effects of motion and susceptibility. In this study, the projection reconstruction technique was further modified by optimizing MR signal frequencies for reconstructing the images. This was done by means of one of two methods. With the first method, a susceptibility map was derived from the raw image data and this map was used to indicate the optimal frequencies for reconstructing the images. The second method of susceptibility correction was a postprocessing technique in which the optimal reconstruction frequencies were selected with use of specific focusing criteria to generate the least blurred image. The effect of using susceptibility map correction on a phantom was demonstrated, and both of these methods were used to improve the visibility of pulmonary structures on images of subjects with normal and abnormal lungs.

Lipid-suppressed single- and multisection proton spectroscopic imaging of the human brain

Spectroscopic images of the brain have great potential in disease diagnosis and treatment monitoring. Unfortunately, interfering lipid signals from subcutaneous fat and poor water suppression due to magnetic field inhomogeneities can make such images difficult to obtain. A pulse sequence that uses inversion recovery for lipid suppression and a spectral-spatial refocusing pulse for water suppression is introduced. In contrast to methods that eliminate fat signal by restricting the excited volume to lie completely within the brain, inversion-recovery techniques allow imaging of an entire section without such restrictions. In addition, the spectral-spatial pulse was designed to provide water suppression insensitive to a reasonable range of B0 and B1 inhomogeneities. Several data processing algorithms have also been developed and used in conjunction with the new pulse sequence to produce metabolite maps covering large volumes of the human brain. Images from single- and multisection studies demonstrate the performance of these techniques.

Twisting radial lines with application to robust magnetic resonance imaging of irregular flow

A problem with magnetic resonance angiograms of vessels containing irregular flow is that flow-induced dephasing can result in voids in the image. These void regions are susceptible to misinterpretation as regions of stenosis or other vessel pathology. Flow-induced dephasing can be minimized by using radial lines to cover k space. However, radial lines provide a very nonuniform, and hence, inefficient, coverage of k space. By twisting the outer portions of the radial trajectories, undistorted images of very rapid and turbulent flow can be obtained with a reasonable number of RF excitations.

Two-dimensional selective adiabatic pulses

Using the technique of separable k-space excitation, we have designed a two-dimensional selective adiabatic pulse that inverts magnetization from a square region in the xy plane with insensitivity to RF variations. We also have designed a two-dimensional adiabatic pulse that inverts selectively in frequency and in one spatial dimension. The pulses should be useful for both MR imaging and spectroscopy. We present experimental results to demonstrate that the two-dimensional adiabatic pulses are feasible on commercial MR imaging systems.

Two-dimensional selective adiabatic pulses

Using the technique of separable k-space excitation, we have designed a two-dimensional selective adiabatic pulse that inverts magnetization from a square region in the xy plane with insensitivity to RF variations. We also have designed a two-dimensional adiabatic pulse that inverts selectively in frequency and in one spatial dimension. The pulses should be useful for both MR imaging and spectroscopy. We present experimental results to demonstrate that the two-dimensional adiabatic pulses are feasible on commercial MR imaging systems.

Effects of RF amplifier distortion on selective excitation and their correction by prewarping

In a magnetic resonance imaging system, an RF power amplifier is employed to boost an RF pulse to sufficient strength to excite the nuclear spins in a subject. The nonideal behavior of this amplifier distorts a selective-excitation pulse, and this distortion in turn degrades the slice profile. We have found two types of nonideal behavior particularly troublesome: nonlinearity and incidental phase modulation. One of their effects is the introduction of an unwanted "skirt" in the out-of-slice region of a slice profile. We present an effective method of correction in which a selective-excitation pulse is prewarped to compensate for the distortion.

Fast angiography using selective inversion recovery

We have developed an enhancement of selective inversion recovery that allows us to obtain high-resolution angiograms in reduced scan time. By applying several read pulses following each tagging inversion pulse, we can obtain several phase encodes in each cardiac cycle, thereby reducing the total scan time required for a complete image. Using this technique, high-resolution angiograms can be obtained in as little as 15 s. Because the phase encodes are collected in short bursts separated by long pauses, care must be taken to maintain uniform signal weighting across phase-encoding views and avoid ghosting. We use an increasing flip-angle sequence to equalize signal level weighting across the readouts. The phase encodes are collected in a special order to minimize ghosting. A postprocessing technique is used to further reduce signal nonuniformity between phase encodes. This fast angiography technique can significantly reduce artifacts due to patient motion during scanning and is especially useful for imaging vasculature in regions of the body where respiratory motion is a problem.

Inhomogeneity correction for in vivo spectroscopy by high-resolution water referencing

One of the most common sources of distortion in in vivo spectroscopy is the inhomogeneity of the main magnetic field. This effect is particularly problematic when performing spectroscopic imaging, as the shim cannot be simultaneously optimized for all voxels. In this paper we present a technique to measure inhomogeneity rapidly, then show how to use the measurement to improve the analysis of the spectrum. This technique can be applied in conjunction with any spectroscopic localization method and any spectral quantitation algorithm. We present results from spectroscopic imaging of phantoms, then show application to a single-voxel water-suppressed proton brain study. We find that the quantitation of the in vivo spectrum is made immune to inhomogeneous line broadening.

Rapid, fully automatic, arbitrary-volume in vivo shimming

MR spectroscopy and many MR imaging methods benefit from a well-shimmed magnet. We have developed a pulse sequence which enables fast and accurate measurement of three-dimensional field maps in vivo, and a data analysis package that allows calculation of shim currents to optimally shim arbitrary selected volumes. A data link to the shim power supply allows automatic update of currents. No intervention by the operator is required. Typical in vivo shimming time is less than 5 min. Performance analysis, phantom, and in vivo results are presented.

Lung parenchyma: projection reconstruction MR imaging

Magnetic resonance (MR) imaging of lung parenchyma is limited by the low proton density and short T2 in the lung as well as the effects of susceptibility and motion. The MR imaging appearance of lung parenchyma was investigated with a pulse sequence that offers some solutions to these problems. This sequence employs projection reconstruction (PR) acquisition gradients and a section-selective excitation pulse designed to eliminate the need to refocus and to allow low-frequency k-space data to be collected with minimal delay. Echo times as short as 50 microseconds can be achieved, producing a proton-density-weighted image. An excised inflated lung specimen and specimens from human subjects with normal lungs (n = 3), pulmonary arteriovenous malformations (n = 1), bronchogenic carcinoma (n = 1), and bullous lung disease with lung metastases (n = 1) were examined. Signal intensity from lung parenchyma and visibility of pulmonary structures were superior on images obtained with the PR MR imaging technique compared with spin-echo images.

1991 I.I. Rabi Award. Estimating oxygen saturation of blood in vivo with MR imaging at 1.5 T

The use of magnetic resonance (MR) imaging is investigated for noninvasively estimating the oxygen saturation of human blood (%HbO2) in vivo by means of relaxation characteristics identified in earlier MR spectrometry studies. To this end, a sequence is presented for determining the T2 of vascular blood in regions in which motions of the body and of the blood itself present a major challenge. With use of this sequence on a commercial 1.5-T whole-body imager, the relationship between the T2 and %HbO2 of blood is calibrated in vitro for the conditions expected in vivo. T2 varies predictably from about 30 to 250 msec as %HbO2 varies from 30% to 96%. T2 values measured in situ for vascular blood in the mediastinum of several healthy subjects qualitatively reflected the behavior observed in vitro. Estimates of %HbO2 for these vessels obtained with the in vitro calibration appear reasonable, particularly for venous blood, although difficulties arise in selecting the appropriate calibration factors. These encouraging initial results support a more systematic study of potential sources of error and an examination of the accuracy of in vivo measurements by comparison with direct measurements of %HbO2 in vessels.

Spectroscopic imaging with multidimensional pulses for excitation: SIMPLE

Proton spectroscopy and spectroscopic imaging in the human brain require the elimination of both water and lipid signals. Strong lipid signals from subcutaneous fat are usually eliminated by confining the excited volume to lie wholly within the skull. Water suppression, however, can be difficult due to both B0 and RF inhomogeneities, which are particularly troublesome in imaging experiments where a relatively large region-of-interest (ROI) is typical. In this paper, we discuss the use of multidimensional selective-excitation pulses (e.g., pulses that are simultaneously selective along two axes) to both define the ROI and provide the necessary water suppression. Pulse sequences providing three-dimensional localization along with water suppression that is insensitive to a range of B0 and RF inhomogeneities are described. Spectra and spectroscopic images (voxel volume = 3.4 cc. acquisition time = 38 min) of various 1H metabolites from a patient with an astrocytoma show clear differences between normal and cancerous tissues and demonstrate the ability of these techniques to be used in vivo.

1H spectroscopic imaging using a spectral-spatial excitation pulse

Excellent water suppression is required to perform in vivo 1H spectroscopic experiments. However water suppression is difficult due to both B0 and RF inhomogeneities. These inhomogeneities are particularly troublesome in spectroscopic imaging experiments where water suppression is required throughout some large region of interest. In this paper, we propose the use of spectral-spatial excitation pulses for such experiments. These two-dimensional pulses are shown to provide water suppression that is insensitive to a range of B0 and RF variations while simultaneously providing spatial localization. Experimental results including images (with voxel volumes ranging from 3.4 to 1.5 cc) of various brain metabolites from both a normal volunteer and a patient with a metastatic lung carcinoma are presented.