Experimental hyaline cartilage lesions: two-dimensional spin-echo versus three-dimensional gradient-echo MR imaging.

The value of magnetic resonance (MR) imaging, with two-dimensional (2D) spin-echo and FISP (fast imaging with steady-state precession) and FLASH (fast low-angle shot) three-dimensional (3D) gradient-echo sequences, for the detection of hyaline cartilage defects of the femoral condyle and the tibial plateau, was investigated in an animal model. In eight dogs, the anterior cruciate ligament was transected in one knee joint, resulting in rapid development of osteoarthritis with degeneration of the hyaline cartilage. At autopsy, 24 cartilage lesions were found, which were classified into four grades. The overall detection of cartilage lesions with MR imaging was poor. Only five of the 24 lesions were visible on 2D spin-echo images, while 11 of 24 were visible on 3D FISP images and 15 of 24 were seen on 3D FLASH images. The best results were obtained in advanced stages of cartilage degeneration, involving ulceration and complete abrasion of the cartilage layer. Signal loss or signal intensity increase in the cartilage layer was seen inconsistently in grades 3 and 4 degeneration. In this animal model, 2D spin-echo imaging was inadequate for the diagnosis of hyaline cartilage lesions, while 3D gradient-echo imaging permitted satisfactory diagnosis in only grade 4 cartilage disease.     

Variation in MR signal intensity across normal human knee cartilage

Signal intensity(SI) of individual pixels on sagittal magnetic resonance (MR)images of normal human knee cartilage was quantified to investigate whether it was related to cartilage proteoglycan content. In five subjects, images were acquired with spin-echo sequences with a TR msec/TE msec ofl,OOO or 7OO/2O and a three-dimensional gradient-echo(GRE)sequence (60/15). In a sixth subject, the GRE sequence alone was used with 15°,30°,and 50° flip angles. In all subjects, SI was maximal in pixel layers of the medial zone and minimal at both cartilage edges, resulting in the presence of a bell-shaped curve of interpixel(zonal) SI variation across the cartilage thickness. The magnitude of SI was dependent on the pulse sequence and flip angle, but the bell shape of the SI variation curve was independent of them. For example, in the medial tibial cartilage, the peak SI was highest with the 1,000/20 spin-echo sequence, intermediate with the 7OO/2O sequence, and lowest with the GRE sequence. The differences were statistically significant. The bell-shaped SI variation curve resembled the curve for zonal variation in cartilage proteoglycan content but not the curves for collagen or free water content. The physiologic basis for this resemblance and the potential usefulness of the findings for early diagnosis of diseases such as osteoarthritis are discussed.     

Delineation of pancreas with MR imaging: Multiobserver comparison of five pulse sequences

The authors compared five magnetic resonance (MR) imaging pulse sequences for their ability to depict the pancreas in 59 patients, each evaluated with at least two of the five sequences. Focal pancreatic carcinomas were present in eight patients. The five sequences were T1-weighted spin echo (T1-SE), fat-suppressed T1-SE (T1-FS). T1-weighted gradient echo (T1-GRE). T2-weighted SE (T2-SE), and T2-weighted fast spin echo (T2-FSE). Using repeated-measures analysis, three blinded observers independently reviewed 198 separate MR imaging series and rated them on a 5-point scale with regard to image quality and depiction of pancreatic borders and the number of sections containing pancreatic and common bile ducts. The most superior and most inferior sections containing pancreas were recorded for each sequence in each patient. The results were compared with analysis of variance, and interobserver agreement was measured with the intraclass correlation coefficient (ICC). For image quality, all sequences were rated good to excellent, with the T1-SE sequence having the highest rating. For clarity of pancreatic borders, however, the T1-FS sequence was rated significantly higher (P<.006) than the other sequences; the T2-SE sequence was least satisfactory. Common bile and pancreatic ducts were seen in the most sections with the T2-FSE sequence. There were no significant differences regarding identification of the most superior and inferior sections containing pancreas, and the ICC was high (.91–.97) for all sequences. For detecting focal carcinomas, no single pulse sequence was sufficient.     

Clinical comparison of three-dimensional MP-RAGE and FLASH techniques for MR imaging of the head.

Three-dimensional (3D) MP-RAGE (magnetization-prepared rapid gradient-echo) imaging was evaluated as a high-resolution 3D T1-weighted brain imaging technique for patients with suspected neurologic disease. Fourteen patients were studied. In five, 3D MP-RAGE images were compared with 3D FLASH (fast low-angle shot) images. Signal difference--to-noise ratios and T1 contrast were not statistically different for 3D MP-RAGE images as opposed to 3D FLASH images. Advantages intrinsic to the application of 3D MP-RAGE sequences include decreased imaging time and decreased motion artifact. With this technique, it is possible to perform a relatively motion-insensitive, T1-weighted screening brain study with voxel resolution of 1.0 x 1.4 x 2.0 mm or smaller, in an imaging time of 5.9 minutes or less--permitting offline (poststudy) reconstruction of high-resolution images in any desired plane.     

Multislab three-dimensional T2-weighted fast spin-echo imaging of the hippocampus: Sequence optimization

Three-dimensional (3D) fast spin-echo (FSE) imaging can produce contiguous thin sections for high-quality multiplanar reconstructions. Such reformatted images may be useful in the evaluation of three-dimensionally complex, curvilinear anatomic structures such as the hippocampus. The authors describe a 3D FSE protocol for T2-weighted Imaging of the hippocampus. The protocol uses an overlapplng-multiple-slab imaging strategy to decrease Imaging times and a modified refocusing radio-frequency pulse train to improve the reformatted images. The authors describe their parameter optimization, discuss the benefits and limitations of the new sequence, and present representative images of healthy volunteers.     

Fast spin-echo imaging of the knee: Factors influencing contrast

Conventional T2-weighted spin-echo magnetic resonance imaging of the knee requires a long TR. Fast spin-echo (FSE) imaging can improve acquisition efficiency severalfold by collecting multiple lines of k space for each TR. Compromises in resolution, section coverage, and contrast inevitably result. The authors examined the compromises encountered in FSE imaging of the knee and discuss the variations in image contrast and resolution due to choices of sequence parameters. For short TR/TE knee imaging, FSE does not appear to offer any advantages, since the increased collection efficiency for one section reduces the available number of sections, so that the total imaging time for a given number of sections remains constant relative to conventional spin-echo imaging. For T2-weighted images, considerable time can be saved and comparable quality images can be obtained. This saved time can be usefully spent on increasing both the resolution of the image and its signal-to-noise ratio, while still reducing total acquisition time by a factor of two. The preferred FSE T2-weighted images were acquired with a TR of 4,500 msec, TE of 120 msec, and eight echoes. The available number of sections is compromised, and the sequence remains sensitive to flow artifacts; however, the FSE sequence appears to be promising for knee imaging.     

Ultrastructural MR imaging techniques of the knee articular cartilage: problems for routine clinical application

The high incidence of cartilage lesions together with new surgical treatment techniques have necessitated the development of noninvasive cartilage evaluation techniques. Although arthroscopy has been the standard for cartilage evaluation, MR imaging has emerged as the imaging method of choice, allowing morphological evaluation of cartilage and cartilage repair tissue, as well as evaluation of its biochemical content. This article deals with current ultrastructural MR imaging techniques for cartilage evaluation, indicating the advantages as well as the drawbacks for routine clinical application.     

MR microscopy of hyaline cartilage: current status

Cartilage degenerative diseases, such as osteoarthritis, affect million of people. Magnetic resonance imaging is presently the most accurate imaging modality in evaluating the state of hyaline cartilage; however, clinical MRI does not accurately reveal early degenerative alterations in cartilage, due mainly to low spatial resolution. Magnetic resonance microscopy (MRM, or microMRI) appears exceptionally well suited to the in vitro or ex vivo study of this heterogeneous tissue, due to its high spatial resolution; however, despite this, further studies are necessary to evaluate the potential of MRM in the detection of early cartilage damage. Herein we briefly review the current applications of MRM in the study of hyaline cartilage. In particular, we review the MR appearance of hyaline cartilage on high-resolution images, the different MRM techniques used to image normal and enzymatically or chemically degraded cartilage and the potential use of contrast agents. The future directions and the relevance of MRM findings for a better understanding of cartilage physiology in health and disease are also discussed.     

Three-dimensional fast spin-echo imaging: Pulse sequence and in vivo image evaluation

Three-dimensional (3D) magnetic resonance imaging allows thin-section acquisition and therefore more accurate multiplanar reconstruction than conventional two-dimensional spin-echo imaging. Unfortunately, addition of a third acquisition plane extends imaging time greatly. With gradient-echo techniques, 3D acquisitions have become clinically useful. These techniques are limited, however, by susceptibility and other field inhomogeneity artifacts and decreased signal-to-noise ratios compared with spin-echo techniques. The authors describe implementation of a true spin-echo 3D technique that, by using fast spin-echo parameters, reduces acquisition time to a clinically useful level. Potential applications of the technique are demonstrated.     

Selective water excitation for faster MR imaging of articular cartilage defects: initial clinical results

Our objective was to compare a water-excitation (WE) 3D fast low-angle shot (FLASH) MR sequence for faster imaging of articular cartilage defects of the knee to a conventional fat-saturated (FS) 3D FLASH MR sequence. This prospective study included 16 knees of 16 patients with suspected cartilage lesions. The MR imaging in transverse and sagittal planes included (a) FS 3D FLASH (TR/TE: 45 ms/11 ms, scan time 8 min, flip angle 50°), and (b) WE 3D FLASH (TR/TE: 28 ms/11 ms, scan time 4 min 58 s, flip angle 40°). For each sequence signal-to-noise ratios (SNR) and contrast-to-noise ratios (CNR) were quantified. The detected cartilage lesions were evaluated using a semi-quantitative four-scale scoring system (grades 0–III). The data were compared between the sequences using the paired Student's t-test. No statistically significant differences between the sequences were found for SNR, CNR, and cartilage defect grading (p=0.14–0.8). The WE 3D FLASH MR imaging seems to be promising for fast imaging of articular cartilage lesions of the knee. Electronic Publication     

Optimization of three-dimensional T1-weighted gradient-echo imaging of the cervical spine.

Various parameters of the three-dimensional (3D) T1-weighted magnetization-prepared rapid acquisition gradient-echo (MP-RAGE) sequence were evaluated to improve spatial resolution while maintaining T1 contrast and a short examination time in imaging of the cervical spine in volunteers. The most dramatic improvements in image resolution occurred by decreasing section thickness to 1.2 mm and increasing the in-plane matrix to 192 x 256, with a 230-mm field of view. The increase in imaging time due to the increased matrix was offset by the elimination of the preparation pulse and wait time, without dramatic changes in contrast-to-noise ratio or overall image quality. Optimum parameters included elimination of the preparation pulse and wait time, 12 degrees flip angle, 192 x 256 matrix, 1.2-mm section thickness, nonselective excitation (coronal acquisition), RF spoiling, and standard k-space ordering, for an examination time of 5 minutes 21 seconds.     

T2-weighted three-dimensional MP-RAGE MR imaging.

The application of three-dimensional (3D) magnetization-prepared rapid-gradient-echo (MP-RAGE) imaging to the acquisition of T2-weighted 3D data sets has been investigated, with a 90 degrees x-180 degrees y-90 degrees-x pulse set (driven equilibrium) for the T2 contrast preparation. A theoretical model was used to study the contrast behavior of brain tissue. The effects of radio-frequency and static-field inhomogeneities and eddy currents on the T2 contrast preparation and the effects of eddy currents on the gradient-echo acquisition resulted in blurring and intensity banding artifacts. With a multistep gradient preparation, these artifacts could be suppressed. With further development, this technique may yield a clinically practical method for obtaining T2-weighted 3D data sets of relatively large volumes (eg, the whole head) suitable for multiplanar reformatting.     

Comparison of breath-hold T1-weighted MR sequences for imaging of the liver

Three rapid T1-weighted gradient-echo techniques for imaging of the liver were compared: fast low-angle shot (FLASH) and section-selective (SSTF) and non-section-selective (NSTF) inversion-recovery TurboFLASH. Ten healthy volunteers were imaged at 1.5 T, with breath-hold images acquired in the transaxial and coronal planes and non-breath-hold images in the transaxial plane. Breath-hold images were evaluated quantitatively and qualitatively, and non–breath-hold images were evaluated qualitatively. FLASH images had significantly higher (P <.001) spleenliver signal difference–to-noise ratios (SD/Ns) than NSTF and SSTF images. Liver signal-to-noise ratios (S/Ns) were significantly higher (P <.001) on FLASH images than on NSTF and SSTF images. With breath hold. FLASH images were rated as having the highest quality in the axial plane, followed by NSTF and SSTF images. In the coronal plane, NSTF images were rated as having the highest quality. For images acquired during patient respiration, NSTF images had the highest quality and showed the least degradation. The results suggest that FLASH images have the highest SD/N and S/N for liver imaging and have the highest quality in the axial plane. In patients who cannot suspend respiration, NSTF images may be least affected by breathing artifact and provide reasonable image quality.     

Minimizing TE in moment-nulled or flow-encoded two- and three-dimensional gradient-echo imaging.

A method for minimizing field-echo delay in moment-nulled gradient-echo imaging is presented. Even though ramps are accounted for, the analysis yields simple closed-form solutions. The method is then generalized to the section-select waveform for three-dimensional volume imaging and to flow encoding for phase-contrast imaging. Three strategies for first-moment selection in phase-contrast imaging are discussed, including a new strategy that always yields the minimum echo delay. Trapezoidal and triangular gradient lobe shapes are analyzed.     

Separation of fat and water in fast spin-echo MR imaging with the three-point dixon technique

A method for suppressing fat in fast spin-echo imaging with the three-point Dixon technique is described. The method differs from the three-point Dixon method used in conventional spin-echo imaging in that the readout gradient instead of a radio-frequency pulse is shifted. This method preserves the Carr-Purcell-Meiboom-Gill nature of the fast spin-echo sequence and hence is less sensitive to magnetic field inhomogeneities and resonance frequency mistiming. As in the original three-point Dixon technique used in conventional spin-echo imaging, three acquisitions are required to estimate the field inhomogeneity and completely separate fat and water. The extra time required is not excessive considering that the fast spin-echo method is frequently applied with multiple signal acquisition. Also, this technique achieves an expected signal-to-noise ratio comparable to 2.67 signal acquisitions, which is approximately 94% of the signal-to-noise ratio obtained with three signal acquisitions. The method is demonstrated with applications to phantoms and a human volunteer.     

T2-weighted MR imaging of the chest: Comparison of electrocardiograph-triggered conventional and turbo spin-echo and nontriggered turbo spin-echo sequences

In 22 patients with a diverse range of thoracic abnormalities, T2-weighted magnetic resonance (MR) images of the chest were obtained with electrocardiograph (ECG)-triggered turbo spin-echo (TSE), ECG-triggered conventional spin-echo (CSE), and nontriggered TSE sequences, and the images were compared. A 5-point rating scale was used by three radiologists experienced in MR imaging of the chest to Independently evaluate the images for (a) freedom from ghosting, (b) clarity of heart wall and cardiac chambers, (c) clarity of mediastinal structures, (d) conspicuity of abnormalities, and (e) overall image quality. Evaluations were analyzed with statistical methods. For freedom from ghosting, clarity of heart wall and cardiac chambers, clarity of mediastinal structures, and overall image quality, the ECG-triggered TSE images were rated higher than the TSE images, which. In turn, were rated higher than the ECG-triggered CSE images at the P=.05 level of significance. No significant differences were seen between the pulse sequences in the conspicuity of abnormalities, although some differences were observed in individual cases. Our results suggest that ECG-triggered TSE imaging provides improved, time-efficient T2-weighted images of the chest.     

Centric phase-encoding order in three-dimensional MP-RAGE sequences: application to abdominal imaging.

Three-dimensional (3D) magnetization-prepared rapid gradient-echo imaging has been proposed as a method for improving signal-to-noise ratio (S/N) and contrast-to-noise ratio (C/N) in rapid abdominal imaging. Originally, a standard sequential phase-encoding order was proposed. In the present study, two approaches to a 3D centric phase-encoding order are presented: (a) application of the two-dimensional (2D) centric order to one of the 3D encoding directions, and (b) an interleaved square spiral order, which is the segmented 3D analog of the 2D centric order. With use of simulation, phantom, and volunteer results, the proposed 3D centric methods are compared in terms of S/N, C/N, and artifacts to the 3D sequential method and 2D magnetization-prepared methods. The second centric approach was found to be superior to the first; however, in general, the 3D technique was found to be inferior to the 2D technique for abdominal imaging because of motion artifact in the 3D image set caused by misregistration among the multiple breath holds required.     

Rapid three-dimensional T1-weighted MR imaging with the MP-RAGE sequence.

The authors investigated the application of three-dimensional (3D) magnetization-prepared rapid gradient-echo (MP-RAGE) imaging to the acquisition of small (32 x 128 x 256) T1-weighted 3D data sets with imaging times of approximately 1 minute. A theoretical model was used to study the contrast behavior of brain tissue. On the basis of these theoretical results, 3D MP-RAGE sequences were implemented on a 1.5-T whole-body imager. Thirty-two-section 3D data sets demonstrating good signal-to-noise ratios and resolution and strong T1-weighted contrast were obtained in 1 minute. Compared with standard short TR/TE spin-echo sequences with the same imaging times and comparable sequence parameters, the 3D MP-RAGE sequence delivered increases of more than 50% in the white matter/gray matter signal difference-to-noise and white matter signal-to-noise ratios, and provided almost twice as many sections. These sequences may find a clinical role in 3D scout imaging and screening and in patients with claustrophobia or trauma.     

Coupled-spin fast spin-echo MR imaging

Distinguishing between lipid and water-containing tissues is clinically important. Current techniques rely on the chemical shift difference between fat and water resonances or differences in relaxation times of the tissues, or a combination of both. A method is presented for separating the signals of lipid protons from those of water protons by using fast spin-echo magnetic resonance imaging based on the principle that lipid protons behave differently from water protons in mul-tiecho sequences. Two images are acquired with different echo train lengths and echo spacing but with identical TEs, and then subtracted to exploit differences in the behavior of lipid and water protons in mul-tiecho sequences. The method is insensitive to B₀ inhomoge-neities or susceptibility effects and provides separate lipid and water images with a high signal-to-noise ratio. The advantages of the method are demonstrated with phantom studies and clinical examples.