Myocardial first-pass perfusion magnetic resonance imaging: a multicenter dose-ranging study.

BACKGROUND: MRI can identify patients with obstructive coronary artery disease by imaging the left ventricular myocardium during a first-pass contrast bolus in the presence and absence of pharmacologically induced myocardial hyperemia. The purpose of this multicenter dose-ranging study was to determine the minimally efficacious dose of gadopentetate dimeglumine injection (Magnevist Injection; Berlex Laboratories) for detecting obstructive coronary artery disease. METHOD AND RESULTS: A total of 99 patients scheduled for coronary artery catheterization as part of their clinical evaluation were enrolled in this study. Patients were randomized to 1 of 3 doses of gadopentate dimeglumine: 0.05, 0.10, or 0.15 mmol/kg. First-pass perfusion imaging was performed during hyperemia (induced by a 4-minute infusion of adenosine at a rate of 140 microg x kg(-1) x min(-1)) and then again in the absence of adenosine with otherwise identical imaging parameters and the same contrast dose. Perfusion defects were evaluated subjectively by 4 blinded reviewers. Receiver-operating curve analysis showed that the areas under the receiver-operating curve were 0.90, 0.72, and 0.83 for the low-, medium-, and high-contrast doses, respectively, compared with quantitative coronary angiography (diameter stenosis > or =70%). For the low-dose group, mean sensitivity was 93+/-0%, mean specificity was 75+/-7%, and mean accuracy was 85+/-3%. CONCLUSIONS: First-pass perfusion MRI is a safe and accurate test for identifying patients with obstructive coronary artery disease. A low dose of 0.05 mmol/kg gadopentetate dimeglumine is at least as efficacious as higher doses.

Gadolinium-enhanced, vessel-tracking, two-dimensional coronary MR angiography: single-dose arterial-phase vs. delayed-phase imaging.

The purposes of our study were to investigate the benefits of using a single dose of an extracellular contrast agent for coronary magnetic resonance angiography (CMRA) and to determine the relative benefits of arterial-phase vs. delayed-phase image acquisition. The right coronary artery was imaged in 10 healthy adults using a breath-hold, two-dimensional fast gradient echo pulse sequence designed for vessel tracking (multiphase, multislice image acquisition). Pre- and postcontrast CMRA was performed. Postcontrast imaging consisted of arterial- and delayed-phase CMRA following a 15 mL bolus (single dose) of contrast media and of delayed-phase imaging following a cumulative 45 mL contrast dose (triple dose). Contrast-enhanced CMRA provided a significantly higher (P < 0.001) signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than noncontrast CMRA. CNR was highest for single-dose arterial-phase CMRA (13.1 +/- 4.5) and triple-dose delayed-phase CMRA (13.0 +/- 4.8), followed by single-dose delayed-phase CMRA (8.4 +/- 3.5) and noncontrast CMRA (4.2 +/- 1.8). Single-dose arterial-phase CMRA provided the best visualization of the distal right coronary artery and was preferred for blinded physician assessments. We concluded that utilization of a single dose of extracellular contrast media improves CMRA, especially if timed for arterial-phase imaging. J. Magn. Reson. Imaging 2001;13:682-689.

First-pass myocardial perfusion MR imaging with interleaved notched saturation: feasibility study.

The authors evaluated a magnetization preparation scheme with a "notched" section profile for T1-weighted first-pass myocardial perfusion magnetic resonance (MR) imaging at 1.5 T. The pulse sequence consisted of a preparation sequence followed by an interleaved gradient-echo echo-planar sequence. Image contrast was evaluated in a feasibility study in 12 adult patients. The notched saturation pulse allowed long magnetization recovery times without sacrificing section coverage. Image contrast between normal and ischemic myocardium was excellent.

Visualization of aortic valve leaflets using black blood MRI.

Although magnetic resonance imaging (MRI) is capable of imaging various physiological parameters associated with the heart valves, it has generally been difficult to visualize the valve leaflets directly. The aortic valve was imaged in 120 patients referred for cardiac MRI to assess myocardial volumes or mass. The average patient age was 37 and ranged from 9 to 75 years. Heart rate ranged from 43 to 100 bpm. Imaging was performed on a 1.5 T scanner equipped with enhanced gradients and a cardiac phased-array coil. A double inversion recovery fast spin-echo sequence was used to acquire short-axis images of the aortic valve in a breath-hold (15 +/- 3 seconds). All three leaflets of the aortic valve were seen in 102 of 120 studies (85%). Two leaflets were detected in another 15 subjects. No leaflets were seen in three individuals. Seven cases of a bicuspid or thickened aortic valves were clearly distinguished from normal valves. The signal-to-noise ratio of aortic leaflets (14 +/- 5) was significantly higher than that of the residual blood signal in the aortic root (7 +/- 4, P < 0.001). MR images showed the aortic valve leaflets in a high fraction of people with suspected normal aortic valves and detected seven cases of abnormal aortic valves. The potential of MRI to study both the anatomic and functional consequences of valvular heart disease warrants further study. J. Magn. Reson. Imaging 1999;10:771-777.

Improved field of view-reducing gradient insert: artifacts and application to cardiac imaging.

An improved homogeneity-spoiling local gradient insert has been constructed and built into the patient bed to reduce its impact on bore size and setup time. It allows the field of view to be reduced without introducing aliasing artifacts. Image quality and artifacts were evaluated for pulse sequences relevant to cardiac imaging. The utility of the insert is demonstrated by a contrast-enhanced perfusion study, in which the reduced field of view allowed a 25% increase in resolution and one more slice to be imaged per heartbeat. J. Magn. Reson. Imaging 1999;10:209-215.

Myocardial velocity gradient imaging by phase contrast MRI with application to regional function in myocardial ischemia.

Velocity-encoded phase contrast magnetic resonance imaging (MRI) has the potential to quantify regional myocardial contractile function with a sensitivity to motion comparable to implanted ultrasonic crystals. An MRI sequence and post-processing algorithm were developed to measure myocardial velocity gradients on a 1.5 T MRI scanner. These methods were validated on a rotating phantom and applied to dogs before (n = 11) and during prolonged coronary occlusion (n = 5). In phantom validation studies, the average absolute error corresponded to motion equivalent to 0.03 +/- 0.04 mm (mean +/- SD) during the repetition time of the experiment. Rigid body corrections during post-processing significantly simplified the interpretation of myocardial velocity vectors. In vivo, rigid body motion contributes substantially to the recorded myocardial velocities in systole and diastole and can give the false impression of regional wall motion abnormalities. After rigid body correction, normal systolic and diastolic velocity vectors in short-axis views of the left ventricle were primarily directed toward the center of the left ventricle. Transmural radial strain rate was 2.0 +/- 0.6 sec-1 during systole and -3.6 +/- 1.1 sec-1 during early diastole in normal canine hearts. Ischemic myocardium was easily discriminated from normal left ventricle by velocity-encoded phase contrast MRI both qualitatively and quantitatively (P < 0.01 in systole and P < 0.05 in early diastole). Although the myocardial velocity images have a spatial resolution on the order of a millimeter, the velocity encoding describes the mechanical consequences of focal myocardial ischemia with sensitivity to submillimeter displacement of the pixels. The three-dimensional nature of velocity-encoded MRI is particularly well suited to the study of the complex motion of the heart in vivo.

Segmented k-space fast cardiac imaging using an echo-train readout.

A segmented k-space fast gradient-echo pulse sequence with an echo-train readout (FGRE-ET) was developed for high-quality cine imaging of the heart in reduced scan times. Using segmented FGRE-ET, cine images of the heart can be acquired in as few as 1-5 heart beats and do not display the geometric distortion and flow-related artifacts typically associated with cardiac echoplanar imaging (EPI). Segmented FGRE-ET was compared with conventional segmented FGRE and with conventional multi-phase EPI in normal volunteers. Segmented FGRE-ET was found to have reduced temporal blurring compared with segmented FGRE for cine imaging in 4 heart beats (P<0.05). Also, segmented FGRE-ET did not display geometric distortion characteristic of conventional EPI (P<0.05). Segmented FGRE-ET may be particularly applicable to functional cardiac stress testing because it allows versatile cine imaging in very short breath-holds.

Reduction of field of view in MRI using a surface-spoiling local gradient insert.

Herein is presented a method for suppressing the magnetic resonance signal to a controlled depth by applying a spatially heterogeneous spoiler field between the slice-select and readout pulses. Eliminating the signal from near-surface regions allows one to shrink the field of view without introducing aliasing artifacts, thereby decreasing imaging time over a smaller defined volume. A unique planar magnetic gradient coil was constructed to generate the spoiler field. Phantom and human subject studies showed that the signal can be suppressed to controlled distances of up to 90 mm from the coil, with modest requirements on power supplies, pulse sequences, and materials, and with no increase in imaging time.

Detection of proton chemical exchange between metabolites and water in biological tissues.

Metabolites in proton chemical exchange with water were detected via the water proton signal using saturation transfer techniques in model systems and biological tissues. The metabolites were selectively saturated and the resulting decrease in the much larger water proton pool was used to monitor the metabolite. This indirect detection scheme can result in a several orders of magnitude increase in sensitivity for metabolites over direct detection methods. A control irradiation scheme was devised to compensate for macromolecular/water magnetization transfer. Using this approach, significant chemical exchange regions at approximately 1 and 2.5 ppm were detected in kidney medulla. Using a difference imaging technique between a control irradiation above (-1.74 ppm) and below (+1.74 ppm) the water resonance, a chemical exchange image of the kidney was calculated. These data revealed a linear gradient of chemical exchange increasing from the cortex to the medulla. Studies on medullary acid extracts and urine revealed that the exchange observed in the kidney was predominantly with low molecular weight metabolites. Urea (1 ppm) was identified as contributing to the kidney/urine chemical exchange; however, other unidentified metabolites may also contribute to this effect. These studies demonstrate that tissue metabolites can be detected and imaged via the water protons using the signal amplification properties of saturation transfer in the presence of water/macromolecule magnetization transfer.

Improved coverage in dynamic contrast-enhanced cardiac MRI using interleaved gradient-echo EPI.

An interleaved gradient-echo echo-planar imaging (IGEPI) sequence was modified for and applied to dynamic contrast-enhanced imaging of the heart. Using IGEPI, images with 3.0 x 3.9 mm nominal in-plane resolution are acquired in 100 ms, enabling eight slices per heartbeat for a heart rate of 60 beats/min. The acquisition speed and use of saturation prepulses allows acquisition of short- and long-axis images during the same contrast bolus. IGEPI maintains the acquisition characteristics required for performing a quantitative first-pass perfusion analysis as well as providing improved coverage compared with conventional fast gradient echo.

Centric ordering is superior to gradient moment nulling for motion artifact reduction in EPI.

Echo-planar imaging (EPI) is sensitive to motion despite its rapid data acquisition rate. Compared with traditional imaging techniques, it is more sensitive to motion or flow in the phase-encode direction, which can cause image artifacts such as ghosting, misregistration, and loss of spatial resolution. Consequently, EPI of dynamic structures (eg, the cardiovascular system) could benefit from methods that eliminate these artifacts. In this paper, two methods of artifact reduction for motion in the phase-encode direction are evaluated. First, the k-space trajectory is evaluated by comparing centric with top-down ordered sequences. Next, velocity gradient moment nulling (GMN) of the phase-encode direction is evaluated for each trajectory. Computer simulations and experiments in flow phantoms and rabbits in vivo show that uncompensated centric ordering produces the highest image quality. This is probably due to a shorter readout duration, which reduces T2* relaxation losses and off-resonance effects, and to the linear geometry of phantoms and vessels, which can obscure centric blurring artifacts.

Using cardiac phase to order reconstruction (CAPTOR): a method to improve diastolic images.

A method is proposed to reconstruct multiphase images that accurately depicts the entire cardiac cycle. A segmented, gradient-recalled-echo sequence (FASTCARD) was modified to acquire data continuously. Images were reconstructed retrospectively by selecting views from each heartbeat based on cardiac phase rather than the time elapsed from the QRS complex. Cardiac phase was calculated using a model that compensates for beat-to-beat heart rate changes. Images collected using cardiac phase to order reconstruction (CAPTOR) depict the entire cardiac cycle and lack the temporal gap that is characteristic of prospectively reconstructed sequences. Time-volume curves of the left ventricle capture the contribution of atrial contraction to end-diastolic volume (EDV). Transmitral phase-contrast flow measurements show a second peak inflow (alpha wave) that is absent in the standard sequence. Because atrial contraction contributes to ventricular EDV, images using CAPTOR potentially may provide a more reliable measure of EDV, stroke volume, and ejection fraction than standard techniques.

The effect of off-resonance radiofrequency pulse saturation on fMRI contrast.

This paper describes the use of off-resonance saturation to further manipulate the blood oxygenation level dependent (BOLD) contrast of fMRI. A customized narrow bandwidth radiofrequency pulse, applied with a range of frequency offsets prior to selection of each slice, was designed and incorporated into a gradient echo EPI sequence. This application takes advantage of the resonance frequency and linewidth differences between the oxygenated and deoxygenated state of blood in human brain during task activation and rest, and is capable of creating an enhancement in the contrast of the BOLD effect. Because of a possible contribution to the signal change from cerebro-spinal fluid, which has a much narrower linewidth and smaller frequency shift compared with the brain tissue, data were also collected using a nulling inversion pulse. The inversion pulse was applied before the off-resonance pulse and data acquisition to eliminate the CSF signal. Functional areas are thus more localized to the brain tissue.

Assessing contrast on MR images.

Magnetic resonance imaging pulse sequences are frequently judged by their ability to facilitate discrimination between pathologic and normal tissue. Objective analysis is usually based on signal intensity measurements. However, the literature shows disagreement as to how this analysis should be performed. The ability to visually differentiate two objects on the basis of signal intensity depends on the contrast-to-noise ratio (CNR). This parameter, however, can be calculated only by measuring the intensity of photons reaching the eye from two distinct objects and, hence, is display dependent. The signal difference-to-noise ratio (SDNR) is a display-independent parameter that reflects the contrast-generating ability of a pulse sequence. When comparing two imaging sequences, the SDNR is proportional to the CNR, assuming the images being compared are displayed so that corresponding regions have the same intensity (i.e., photon fluxes). Because the SDNR is display independent, it should be the preferred parameter for assessing the contrast-generating ability of a pulse sequence. The value and limitations of these parameters are discussed.

A method to improve the B0 homogeneity of the heart in vivo.

A homogeneous static (B0) magnetic field is required for many NMR experiments such as echo planar imaging, localized spectroscopy, and spiral scan imaging. Although semi-automated techniques have been described to improve the B0 field homogeneity, none has been applied to the in vivo heart. The acquisition of cardiac field maps is complicated by motion, blood flow, and chemical shift artifact from epicardial fat. To overcome these problems, an ungated three-dimensional (3D) chemical shift image (CSI) was collected to generate a time and motion-averaged B0 field map. B0 heterogeneity in the heart was minimized by using a previous algorithm that solves for the optimal shim coil currents for an input field map, using up to third-order current-bounded shims (1). The method improved the B0 homogenelty of the heart in all 11 normal volunteers studied. After application of the algorithm to the unshimmed cardiac field maps, the standard deviation of proton frequency decreased by 43%, the magnitude 1H spectral linewidth decreased by 24%, and the peak-peak gradient decreased by 35%. Simulations of the high-order (second- and third-order) shims in B0 field correction of the heart show that high order shims are important, resulting for nearly half of the improvement in homogeneity for several subjects. The T2* of the left ventricular anterior wall before and after field correction was determined at 4.0 Tesis. Finally, results show that cardiac shimming is of benefit in cardiac 31P NMR spectroscopy and cardiac echo planar imaging.

MR imaging contrast in human brain tissue: assessment and optimization at 4 T.

PURPOSE: To evaluate multiple magnetic resonance (MR) imaging sequences for their ability to provide T1 and T2-weighted images at a field strength of 4 T, and to validate previously obtained relaxation time measurements. MATERIALS AND METHODS: Different spin-echo, inversion-recovery (IR), gradient-recalled acquisition in the steady state (GRASS), and magnetization transfer contrast-enhanced GRASS sequences were evaluated in a single section, each in at least four volunteers. Also, interleaved echo-planar imaging (IEPI) and interleaved gradient-recalled echo (IGRE) sequences were analyzed and compared with standard sequences. RESULTS: Predicted contrast behavior, according to MR relaxation time measurements, was found to agree well with that of the validation experiments. Fair T1 contrast can be achieved on MR images at 4 T, contrary to early predictions. Under other conditions, however, such as partially spin-density-weighted parameters, lower contrast is observed at the high field strength when compared with conventional field strengths. CONCLUSION: The longer T1 values at higher field strengths have a substantial effect on image contrast. Depending on the sequence and parameters chosen, good image contrast can be realized at 4 T. IEPI and hybrid IR IGRE sequences are useful in shortening the prolonged examination times owing to the longer T1 values at higher field strengths.

Thoracic Kaposi sarcoma in AIDS: CT findings.

Cases of acquired immunodeficiency syndrome (AIDS) related Kaposi sarcoma (KS) were reviewed to characterize the spectrum of thoracic findings seen with chest CT. Of 15 patients with AIDs-related KS involving the chest, 13 (87%) demonstrated pulmonary parenchymal disease characterized by multiple, bilateral flame-shaped or nodular lesions with ill-defined margins distributed along bronchovascular bundles. Pleural disease was noted in 10 (67%) patients, characterized by 9 cases of pleural effusions and 1 case of pleural implants. Chest wall disease involving the sternum, ribs, thoracic spine, and/or subcutaneous tissue was noted in eight (53%) patients. Although pleural and parenchymal lung disease are recognized manifestations of thoracic KS, there is also a high incidence of extrapulmonary chest disease evident on CT in patients with AIDS.

Magnetization transfer contrast: MR imaging of the knee.

The use of magnetization transfer contrast (MTC) in magnetic resonance imaging of the human knee was evaluated in this study. MTC is generated by irradiating the macromolecular protons in tissue with a low power off-resonance radio-frequency field. This results in a decrease in water proton signal intensity where a tight magnetic coupling between water and macromolecules exists. With this approach, the authors have demonstrated that MTC can improve contrast in standard single-section, gradient-recalled-echo images of the knee with regard to fat-muscle and cartilage-synovial fluid comparisons. The effect of changes in repetition time, echo time, and flip angle were also quantitatively evaluated. More important, MTC was shown to generate useful cartilage-synovial fluid contrast on high-resolution three-dimensional images, in which contrast is difficult to generate. This approach may not only provide better structural information about the knee, but may also provide noninvasive insight into the structure and biochemical composition of cartilage in vivo.