Pulsation artifact in short TR MR imaging and angiography: Exacerbation with signal averaging

Averaging the signals from more than one excitation per phase-encoding view increases the signal-to-noise ratio and, in conventional spin-echo magnetic resonance imaging, reduces most motion artifacts. To determine the effects of signal averaging on two-dimensional gradient-echo images, acquisitions with different TRs and with no averaging versus multiple-signal averaging were compared in a pulsatile flow phantom and the human abdominal aorta. Intraview (each view repeated before changing the phase-encoding value) and interview (obtaining all views sequentially and then repeating the entire set) averaging methods were used. Pulsation artifacts were present on all images of the flow phantom and the aorta. Intraview signal averaging, the method most commonly used, exacerbated rather than ameliorated pulsation artifacts with short TR sequences. Pulsation artifacts on two-dimensional images obtained with a short TR can be minimized by completing the acquisition as rapidly as possible, avoiding signal averaging. If signal averaging is used for short TR images, it should be interview averaging.     

Recurrent thyroid carcinoma: characteristics on MR images

Magnetic resonance (MR) imaging was used in 32 patients, including eight with benign disease, after partial or total thyroidectomy to determine sensitivity and specificity of MR imaging in the detection of tumor recurrence, to compare signal intensities of scar versus recurrent tumor qualitatively and quantitatively, and to define the extent of recurrent tumor. Findings from surgery (n = 23), needle biopsy (n = 1), or clinical follow-up (n = 8) were used for verification. Of 24 patients with primary thyroid carcinoma, 15 had recurrence and nine had a normal postsurgical thyroid bed. Diagnosis from MR images was correct in 20 cases, but false positive in three and false negative in one. Local recurrence was characterized by low to medium intensity on short repetition time (TR)/short echo time (TE) images and medium to high intensity on long TR/long TE images. Scar in the normal postsurgical thyroid bed showed low intensity on both short and long TR/TE images. Local recurrence of thyroid carcinoma and lymph node metastasis produced positive contrast compared with muscle on short TR/short TE (31 + 19%) and long TR/long TE (85 + 30%) images; fibrosis produced negative contrast, particularly on long TR/long TE (-56, -80%) images. These results indicate the capability of MR imaging in the evaluation of recurrence of thyroid tumors and in the differentiation of abnormal tissue due to tumor recurrence from postoperative fibrosis by means of signal contrast relative to a reference tissue.     

MR signal intensity of parathyroid adenomas: correlation with histopathology

Recent experience has shown that parathyroid adenomas vary in their MR signal intensity, which raises the question of whether the signal intensity is related to different histologic characteristics. In order to address this question, 10 patients who had MR imaging studies (four at 0.35 T, six at 1.5 T) showing large- to medium-sized parathyroid adenomas and who subsequently underwent surgery with histologic proof of the lesion were evaluated. The MR appearance was compared with histologic characteristics. The adenomas were classified into three groups according to the MR appearance: group I, low signal intensity on short TR/TE images, high signal intensity on long TR/TE images (n = 5); group II, low signal intensity on short and long TR/TE images (n = 3); group III, high signal intensity on short and long TR/TE images (n = 2). Histologic analysis revealed that the major features of each group were different. High cellularity without degeneration or fibrosis was observed for all five adenomas from group I. In group II, all three adenomas showed cellular degenerative changes, old hemorrhage with hemosiderin-loaded macrophages, and/or fibrosis. In group III, both adenomas showed evidence of acute hemorrhage without significant degenerative or fibrotic changes. These data suggest that the signal intensity of parathyroid adenomas on T1- and T2-weighted images corresponds at least in part to differences in histologic composition.     

MR imaging artifacts of the axial internal anatomy of the cervical spinal cord

Transverse scans of the spinal cord routinely demonstrate signal variations related to the internal anatomy of the cord that do not accurately conform to histologic cross sections. This study evaluates the MR appearance of the axial anatomy of the spinal cord and provides correlation to histologic sections as a means to understand this discordance so that disease can be recognized more readily. Short TR/TE spin-echo studies, cardiac-gated multiecho spin-echo studies, and gradient-refocused-echo studies of normal excised human spinal cords, a normal volunteer, and gelatin phantoms were obtained by using the same imaging parameters at 1.5 T. Imaging artifacts were further investigated by using both a 128 x 256 and 256 x 256 matrix with a varying phase-encoded axis. Histologic sections of the excised cords, which were stained for myelin, iron, and cell bodies (Nissl), were used for correlation to the images. We found that significant Fourier truncation and partial-volume imaging artifacts modulated the MR display of the cord. On short TR/TE images a ring of high signal at the periphery of the cord was due to a truncation artifact. The appearance of the central portions of the gray and white matter was affected variably by partial-volume averaging depending on the matrix size. White-matter tracts of the cord were always lower in signal than was the gray matter on all pulse sequences. This finding was not due to iron deposition or CSF motion artifacts. We suspect that this probably was related to dense, longitudinal organization of spinal tracts and resultant anisotropy of water molecule motion similar to that seen in the pyramidal tracts, tendons, and ligaments. We recommend the use of a 128 x 256 matrix with two averages (four excitations) when obtaining axial scans of the spinal cord in living subjects. Although truncation artifacts diminish image quality, the quality is superior to that of images obtained with a 256 x 256 matrix, in which longer scanning times result in motion artifacts and reduced signal to noise.     

Fat-saturation MR imaging of the upper abdomen

The fat-saturation (fatsat) MR technique decreases the signal intensity of fat, thereby enhancing the definition of upper abdominal organs and reducing artifacts while maintaining the T1 and T2 information available on spin-echo sequences. To evaluate the potential of fatsat in examining the abdomen, we conducted a prospective study involving 30 subjects, including four normal volunteers, 18 patients investigated for liver disease, and eight patients studied for miscellaneous abdominal disease. Short TR, 300-600/15-20 (TR/TE), and long TR, 2000-2500/20-30, 70-80, spin-echo images with and without fatsat were compared. The images were evaluated both qualitatively and quantitatively. Qualitative assessment was made with receiver-operating-characteristic (ROC) curve analysis of the confidence level of observers to detect the presence of disease, comparing fatsat with standard spin-echo sequences. ROC analysis showed greater interpreter confidence and accuracy for fatsat sequences than for standard spin-echo sequences. The measured signal-difference-to-noise (SD/N) ratio comparing upper abdominal organs with surrounding tissue revealed the highest values for short TR/TE regular spin echo, followed by short TR/TE fatsat. The highest SD/N ratio for hepatic masses was with long TR/TE fatsat followed by short TR/TE fatsat. The results of this study suggest that the fatsat technique may improve abdominal MR imaging.     

Optimization of the technic in the study of the upper abdomen using magnetic resonance. I. Artifact reduction

Complete motion artifact suppression is possible in abdominal MR imaging with the simple optimization of sequence parameters, with no need for special softwares. The authors have studied the influence of sequence parameters modification on the signal/noise relation and on the presence of motion artifacts. The tested parameters included Repetition Time (TR), 150 to 2000 ms, Echo Time (TE), 20 to 120 ms, and the number of acquisitions, 2 to 16. In T1-weighted sequences, the major advantages were offered by the short TR and short TE association, with many acquisitions. Optimal signal/noise relation and complete motion artifact suppression were thus obtained. In T2-weighted sequences, TE values had to be changed according to the desired contrast enhancement, and the number of acquisitions could not be increased to more than 2 in order to keep the acquisition time short.     

Detection of hepatic metastases: analysis of pulse sequence performance in MR imaging

Forty-three patients with liver metastases were imaged using 14 different pulse sequences (average, 7.5 sequences per patient) to allow direct comparison of their performance. "T2-weighted" spin-echo (SE) images, "T1-weighted" inversion recovery (IR) images, and "T1-weighted" SE images were obtained using a wide range of timing parameters. Pulse sequence performance was quantitated by measuring liver signal-to-noise (S/N) ratios and cancer-liver signal difference-to-noise (SD/N) ratios. Data were standardized to reflect a constant imaging time of 9 minutes for all pulse sequences. The SE 2,000/120 (TR [repetition time]/TE [echo time]) sequence resulted in the greatest SD/N ratio of the T2-weighted SE sequences but also yielded the low S/N ratios, poor anatomic resolution, and motion artifacts common to all T2-weighted SE images. IR sequence images were also sensitive to motion artifacts because of the use of a long TR (1,500 msec). Short TR/TE T1-weighted SE sequences (SE 260/18) had the greatest SD/N ratio (P less than .05), S/N ratio, and anatomic resolution. Furthermore, extensive signal averaging appears to be a powerful solution to all types of motion artifacts in the abdomen.     

Chondrosarcomas of the skull base: MR imaging features.

The magnetic resonance (MR) images from 17 patients with chondrosarcomas of the skull base were retrospectively reviewed to characterize the size, location, signal intensity, and extension of these tumors. Eleven patients with chondrosarcomas received intravenously administered gadopentetate dimeglumine. In 16 patients, computed tomographic (CT) scans were obtained to evaluate intratumorous mineralization and bone erosion. On short repetition time (TR)/echo time (TE) MR images, chondrosarcomas generally had low to intermediate signal intensity; on long TR/TE MR images, they generally had very high signal intensity. Signal heterogeneity on long TR/TE MR images was seen in 10 of 17 tumors (59%) and was caused by matrix mineralization, fibrocartilaginous elements, or both. Matrix mineralization was demonstrated with CT in seven of the 16 chondrosarcomas. Chondrosarcomas showed marked enhancement after administration of gadopentetate dimeglumine in either a heterogeneous (n = 8) or homogeneous (n = 3) pattern. The information about the size and extent of these neoplasms was important in the choice of surgical approaches for gross total resection of tumor.     

Multiple spin echo magnetic resonance imaging of the brain

Advantages of 3D multiecho: 1. High signal to noise ratio which is useful for: a) Long TE b) Short TR c) Thin slices 2. Reduced paradoxical enhancement of blood vessels 3. Contiguous slices 4. Number of slices per scan is not reduced by short TR or long TE 5. Thin slices easily generated 6. A short TR multiecho sequence can produce a spectrum of images reflecting a range of both T1 and T2 weighting in the same scan. This information can often improve specificity. Disadvantages of 3D multiecho: 1. Increased motion sensitivity of 3D acquisitions 2. Scan times increase with TR.     

Musculoskeletal MR imaging: turbo (fast) spin-echo versus conventional spin-echo and gradient-echo imaging at 0.5 tesla

The value of T2-weighted fast spin-echo imaging of the musculoskeletal system was assessed in 22 patients with various neoplastic, inflammatory, and traumatic disorders. Images were acquired with high echo number (i.e., echo train length) fast spin-echo (FSE; TR 2000 ms, effective TE 100 ms, echo number 13, lineark-space ordering), conventional spin-echo (SE; TR 2000 ms, TE 100 ms) and gradient-echo (GRE) sequences (TR 600 ms, TE 34 ms, flip angle 25°). Signal intensities, signal-to-noise ratios, contrast, contrast-to-noise ratios, lesion conspicuousness, detail perceptibility, and sensitivity towards image artifacts were compared. The high signal intensity of fat on FSE images resulted in a slightly inferior lesion-to-fat contrast on FSE images. However, on the basis of lesion conspicuity, FSE is able to replace time-consuming conventional T2-weighted SE imaging in musculoskeletal MRI. In contrast, GRE images frequently showed superior lesion conspicuity. One minor disadvantage of FSE in our study was the frequent deterioration of image quality by blurring, black band, and rippling artifacts. Some of these artifacts, however, can be prevented using short echo trains and/or short echo spacings.     

Conventional and fast spin-echo MR imaging: Minimizing echo time

Magnetic resonance imaging is frequently complicated by the presence of motion and susceptibility gradients. Also, some biologic tissues have short T2s. These problems are particularly troublesome in fast spin-echo (FSE) imaging, in which T2 decay and motion between echoes result in image blurring and ghost artifacts. The authors reduced TE in conventional spin-echo (SE) imaging to 5 msec and echo spacing (E-space) in FSE imaging to 6 msec. All magnetic gradients (except readout) were kept at a maximum, with data sampling as fast as 125 kHz and only ramp waveforms used. Truncated sine radio-frequency pulses and asymmetric echo sampling were also used in SE imaging. Short TE (5.8 msec) SE images of the upper abdomen were compared with conventional SE images (TE =11 msec). Also, FSE images with short E-space were compared with conventional FSE images in multiple body sites. Short TE significantly improved the liver-spleen contrast-to-total noise ratio (C/N) (7.9 vs 4.1, n = 9, P <.01) on T1-weighted SE images, reduced the intensity of ghost artifacts (by 34%, P <.02), and increased the number of available imaging planes by 30%. It also improved delineation of cranial nerves and reduced susceptibility artifacts. On short E-space FSE images, spine, lung, upper abdomen, and musculoskeletal tissues appeared crisper and measured spleen-liver C/N increased significantly (6.9 vs 4.0, n = 12, P <.01). The delineation of tissues with short T2 (eg, cartilage) and motion artifact suppression were also improved. Short TE methods can improve image quality in both SE and FSE imaging and merit further clinical evaluation.     

Cardiac MRI: comparison between single-shot fast spin echo and conventional spin echo sequences in the morphological evaluation of the ventricles

PURPOSE: Black blood single shot FSE sequences (Nffse) employ 180 degrees RF refocalisation pulses preceded by an inversion RF double pulse associated to presaturation pulses. The latter produce signal void of the external volume, and possible reduction of the field of view without wrap-around artifacts along the phase coding direction. The aim of our study was to compare the diagnostic possibilities of the Nffse sequences with those of conventional SE study of cardiac morphology. MATERIAL AND METHODS: Twenty-five patients (19 males and 9 females with age ranging from 20 to 54 years) presented findings suggesting right ventricular arrhythmogenic dysplasia. MR examinations were performed with a 1,5 T unit (GE Signa Horizon Echospeed 8.3, Milwaukee, USA) and Torso Phased Array coil positioned at thoracic level. The morphologic study was performed with SE multiphase-multislice ECG-gated sequences (TR: R-R, TE: 30 ms, FOV 320X250, matrix 160X256, slice thickness 10 mm, acquisition time about 5 minutes) and Single-Shot FSE Half Fourier sequences (TR: R-R, TE: 30 ms, flip angle 120 degrees, ETL 30-40, FOV 360X180, Phase FOV 0,5, VBW 64 MHz, slice tickness 10 mm, acquisition time about 10-12 seconds), by imaging along the long and short axis. The study was completed with Fast Gradient Echo sequences (TR: 9ms, TE: 8,2ms, flip angle 25 degrees, VBW 15,63 MHz, FOV 320X250, 10 mm slice thickness, matrix 128X256), subsequently assessed by cine-MR. In order to compare both sequences, two experienced radiologists performed an analysis of quantitative parameters (signal intensity ratio between fat and muscular interventricular septum) and qualitative parameters (double blind evaluation for the presence of cardiac and respiratory artifacts). RESULTS: The signal intensity ratio for the Nffse sequence images was 4.63 +/- 1.56 on the long axis and 7.69 +/- 2.46 on the short axis, whereas it was 3.17 +/- 0.64 on the long axis and 3,50 +/- 0,75 on the axis one for SE images, with a statistically significant difference (p<0,001 and p<0.002 for the long and short axis, respectively). The two radiologists evaluation of the magnitude of artifacts on the SE and Nffse images was similar only as regards the images with significant artefacts alone. Nffse images consistently afforded a detailed evaluation of the right ventricular wall, although blurring artifacts were more common than with good quality SE images. Presence of fatty infiltration of the right ventricle wall was observed in 5 out of 25 patients. In the remaining 20 patients no fatty substitution of the muscular wall of the right ventricle was observed. DISCUSSION AND CONCLUSIONS: The Nffse sequences provide a number of gated multiphase-multislice images, similar to that obtained by conventional SE sequences, in one breath-hold time interval. Due to high intrinsic contrast and reduction of motion artifacts, the Nffse sequences allow a good evaluation of the ventricular morphology and subepicardial and paracardiac adipose tissue. Image quality can be suboptimal due to blurring artifacts. Therefore Nffse sequences can be advantageously employed to image patients with suspected right ventricular arrhythmogenic dysplasia, whenever conventional SE images exhibit substandard quality.     

Femoral head avascular necrosis: correlation of MR imaging, radiographic staging, radionuclide imaging, and clinical findings

To better correlate the appearance of avascular necrosis (AVN) of the femoral head on magnetic resonance (MR) images with the stage of disease, MR images of 56 proved AVN lesions were compared with staging from corresponding radiographs (n = 56), Tc-99m scans (n = 41), and grade of symptoms (n = 28). Fractures complicating AVN were seen in 28 (50%) of 56 radiographs (radiographic stages III-V). With long repetition (TR) and echo delay (TE) times, a characteristic "double line sign" consisting of high signal intensity inside a low-intensity peripheral rim was seen in 45 lesions (80%). The central region within the rim was isointense with marrow fat on both short and long TR and TE images in 20 (71%) of 28 lesions uncomplicated by fracture (stages I-II) but in only four (14%) of 28 stage III-V lesions (P less than .001). Symptoms were least severe in lesions isointense with fat and most severe in lesions with low-signal central regions at short and long TRs and TEs. The peripheral double line sign on long TR/TE images may add specificity to the diagnosis of AVN by MR imaging. A chronologic pattern of central MR signal features is presented which may allow staging of AVN by MR imaging.     

High-field MR imaging of extracranial hematomas

The MR features of 20 extracranial hematomas studied on a 1.5-T system and imaged with both short repetition-time/echo-time (TR/TE) and long TR/TE pulse sequences were reviewed. In four of five acute hematomas (those less than 7 days of age), signal intensity was markedly decreased on long TR/TE images and was either intermediate or slightly decreased on short TR/TR images. Fourteen subacute hematomas (7 days to 7 weeks of age) and one chronic hematoma (9 months) were studied. The appearance of the subacute lesions varied from intermediate to high intensity on short TR/TE sequences, but all demonstrated increased signal on long TR/TE sequences. A low-signal rim was noted at the margin of nine subacute lesions. In one patient with this finding, pathologic examination showed that the low-signal margin corresponded to a region containing hemosiderin-laden macrophages at the periphery of the hematoma. These results correlate well with those reported for intracranial hematomas examined at this field strength. We conclude that analysis of signal-intensity patterns at 1.5 T is useful in staging the evolution of hematomas.     

Preliminary clinical results with low flip angle spin-echo MR imaging of the head and neck

A new approach for producing primarily T2- and proton-density-weighted MR images in less time than the conventional long TR, long TE imaging is to reduce the TR of a double spin-echo pulse sequence and to also reduce the RF excitation flip angle to minimize the resulting T1 sensitivity. In preliminary studies with a human volunteer and five patients with various diseases of the head and neck, conventional long TR, long TE and short TR, short TE images were compared with short TR, long TE images with reduced flip angles (45 degrees, 30 degrees), which required only 40% of the imaging time of the long TR images. The latter images showed a similar contrast pattern to the conventional T2-weighted image, and contrast-to-noise measurements indicated an increase in contrast between the lesion and nearby tissue when the flip angle was reduced. Furthermore, the maximum contrast/noise per unit imaging time on the short TR, long TE image was comparable to that on the long TR, long TE image. Optimization of the flip angle with short TR allows a substantial reduction in imaging time but with a reduction in multislice capability. This technique will be most useful in areas of complex anatomy where two or more orthogonal imaging planes are required, such as the head and neck.     

MR imaging of intracranial metastatic melanoma

Ten patients with intracerebral metastases from malignant melanoma were evaluated with magnetic resonance (MR) imaging performed at 1.5 T using spin-echo techniques. On the basis of histopathologic findings in three of 10 cases and CT appearances in all 10 cases, three patterns were identified on analysis of MR signal intensities in both short repetition time/echo time (TR/TE) and long TR/TE spin-echo scans. In comparison to normal cortex, nonhemorrhagic melanotic melanoma appeared markedly hyperintense on short TR/TE images and isointense, mildly hypointense on long TR/TE images. Nonhemorrhagic, amelanotic melanoma appeared isointense or mildly hypointense on short TR/TE and isointense or mildly hyperintense on long TR/TE images. Hemorrhagic melanoma varied in appearance, depending on the stage of hemorrhage. Melanotic, nonhemorrhagic melanoma can be distinguished from early and late subacute hemorrhage by its signal intensity on long TR/TE images. Spin-echo MR appears to be the method of choice for diagnosing melanotic metastases.     

The magnetic resonance imaging appearance at 1.5 Tesla of cartilaginous tumors involving the epiphysis

Three cases of lytic, calcified epiphyseal lesions with plain film and computed tomography features suggestive of chondroblastoma were imaged by magnetic resonance imaging. Histopathologic correlation was obtained in each case. Two cases of chondroblastoma showed low signal intensity on both short (TR600/TE20ms) and long (TR2500/TE80ms) spin echo (SE) images. The third case, a clear cell chondrosarcoma, demonstrated increased signal intensity on moderately T2 weighted (TR2500/TE40ms) images. These findings suggest that magnetic resonance imaging may be helpful in distinguishing these lesions.     

Painful sickle cell crisis: bone marrow patterns observed with MR imaging

Eleven patients with homozygous sickle cell anemia (SCA) undergoing painful crisis were studied with magnetic resonance (MR) imaging. The signal intensity of bone marrow was diffusely decreased in the axial and peripheral skeleton on short repetition time (TR)/echo time (TE) images and long TR/TE images, which suggested hematopoietic marrow hyperplasia and was confirmed by isotope marrow scans in five patients. Focal areas of further decrease in signal intensity were seen on short TR/TE images in 12 of the 14 (86%) painful joints and three of the five (60%) painless joints. In the painful joints, these focal areas converted to high signal intensity on long TR/TE images, presumably due to edema, which suggested acute marrow infarction. In the painless joints, these low-intensity focal areas remained as low signal on long TR/TE images, which suggested absence of edema and thus areas of old infarction or fibrosis. These results indicate that MR imaging may enable differentiation between acute and chronic marrow infarcts in patients with SCA and serve as a useful guide in monitoring and directing therapy.     

3.0 Tesla magnetic resonance angiography of endovascular aortic stent grafts: phantom measurements in comparison with 1.5 Tesla

RATIONALE AND OBJECTIVES: This study evaluated different stent grafts by 3 T magnetic resonance angiography (MRA) with respect to lumen visibility, susceptibility-induced signal loss, and type of stent artifacts compared with 1.5 T MRA in a phantom model. METHODS: Six different stent-grafts (tube: n = 3, bifurcated: n = 3) were evaluated by 3 T and 1.5 T MRA using a tube phantom. MRA was performed using T1-weighted sequences at both systems with comparable parameters (3T: TR 5.4/TE 2.0/FA 30 degrees, 1.5 T: TR 6.2/TE 2.2/FA 30 degrees). A blind study of the image quality, including artifacts, was performed by 3 radiologists. Furthermore, signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) values were calculated. Statistical analysis was performed with Student's t test (P < 0.05). RESULTS: One Elgiloy stent graft showed almost a complete intraluminal signal loss at 1.5 and 3 T. All other models could be evaluated by both systems by MRA, resulting in a favorable lumen visibility (score: 1) for prostheses made of nitinol. Scores for overall image quality and artifacts were the same for both MR systems. SNR and CNR values of the stented part of the vessel phantom increased from 320 +/- 33 to 618 +/- 40 and from 306 +/- 34 to 596 +/- 40 at 3 T when compared with 1.5 T, resulting in a significant signal gain of 93% at the higher field strength. CONCLUSIONS: 3 Tesla MRA of aortic stent grafts in a phantom model demonstrates an increase in SNR and CNR when compared with 1.5 T. However, the magnitude of imaging artifacts as well as coherent intraluminal signal loss within the stent does not increase equally in both MR systems.     

MR imaging of focal nodular hyperplasia of the liver

The magnetic resonance appearance of three lesions of focal nodular hyperplasia observed in two patients is reviewed. All three lesions demonstrated isointensity with the liver on the short repetition time (TR) and echo time (TE) spin echo images. One lesion also showed a central area of low signal intensity that pathologically corresponded to a scar. The long TR and TE images had a varied appearance, from mixed high signal intensity to predominantly isointense with the liver. The histopathology for each lesion was carefully reviewed. The primary microscopic difference between the lesion that was predominantly isointense with liver on the long time TR and TE images and the other two, which were of mixed high signal intensity on the long time TR and TE images, was the former's relative lack of fibromuscular obliteration of septal blood vessels in the scar. Focal nodular hyperplasia can have at least two possible appearances at 1.5 T; in these three lesions there was consistent isointensity with the liver on the short TR and TE images.