Diffusion tensor MR imaging and fiber tractography: technical considerations

This second article of the 2-part review builds on the theoretic background provided by the first article to cover the major technical factors that affect image quality in diffusion imaging, including the acquisition sequence, magnet field strength, gradient amplitude, and slew rate as well as multichannel radio-frequency coils and parallel imaging. The sources of many common diffusion image artifacts are also explored in detail. The emphasis is on optimizing these technical factors for state-of-the-art diffusion-weighted imaging and diffusion tensor imaging (DTI) based on the best available evidence in the literature. An overview of current methods for quantitative analysis of DTI data and fiber tractography in clinical research is also provided.     

Dynamic susceptibility contrast MR imaging: correlation of signal intensity changes with cerebral blood volume measurements

Cerebral blood volume (CBV) maps derived from dynamic susceptibility contrast (DSC) magnetic resonance (MR) imaging provide valuable information regarding intracranial micro-hemodynamics and have been helpful in characterizing primary brain tumors and guiding stereotactic biopsy. Another parameter, the maximum signal drop (MSD) during the first pass of intravascular contrast bolus due to T2* effect, can also be measured directly without extensive post-processing and data manipulation. The purpose of our study is to determine whether MSD maps provide information similar to CBV maps in patients presenting with intracranial mass lesions. Twenty-nine patients with various intracranial mass lesions were studied with DSC MR imaging prior to stereotactic biopsy or volumetric resection. Maps of both CBV and MSD are calculated on a pixel-by-pixel basis and displayed as color overlays over the raw images. Relative CBV (rCBV) and MSD (rMSD) values were measured in regions of interest (ROIs) within areas of abnormality and compared. In addition, computer-generated noise was added to the data to estimate the sensitivity of each measurement to noise. The rMSD values were strongly correlated with rCBV values (r = 0.87, P = 0.0001). CBV values were much more sensitive to added noise than MSD values (P < 0.01). MSD maps derived from DSC MR imaging provide information similar to CBV maps in patients with intracranial mass lesions. MSD maps are a simple and reliable indicator of vascularity that can easily be incorporated into routine MR imaging.     

Comparative signal intensity measurements in dynamic gadolinium-enhanced MR mammography

Increases in signal intensity enhancement were measured in defined regions of interest (ROIs) to allow distinction between malignant and benign tumors with dynamic gadolinium-enhanced magnetic resonance (MR) mammography. Twenty patients with palpable breast lesions (15 malignant, five benign) underwent MR mammography. The dynamic gradient-echo sequence was performed with intravenous bolus injection of gadopentetate dimeglumine and consisted of 25 images with a time resolution of 30 seconds. Contrast enhancement was calculated by comparing user-defined ROIs on pre- and postcontrast images. An increase in signal intensity of 70% or more on the 1-minute postcontrast image was used as the criterion of malignancy. MR mammographic results correlated with histopathologic findings in all patients when the defined ROI was in the most enhancing part of the tumor. For the ROI in areas of submaximal enhancement or when the ROI surrounded the whole lesion, only five and nine tumors, respectively, fulfilled the malignancy criterion. All malignant tumors showed large variations in signal intensity enhancement that depended on the position of the ROI in the tumor. Dynamic, gadolinium-enhanced MR mammography allows distinction of benign from malignant breast tumors when the selected ROI is in the most enhancing part of the lesion.     

Unimpinged and impinged anterior cruciate ligament grafts: MR signal intensity measurements

Regionalized magnetic resonance (MR) signal intensities were quantitatively measured in impinged and unimpinged anterior cruciate ligament (ACL) grafts. Images were obtained with a 1.5-T imager, and signal intensity was measured in the proximal, middle, and distal thirds of the graft. In 15 unimpinged ACL grafts, the signal intensity remained low and did not vary during the 1st year of graft implantation (45 images). In contrast, 17 impinged ACL grafts showed an increase in signal intensity in the distal two-thirds of the graft that persisted 1-3 years after implantation (P less than .001). Unimpinged grafts were placed in tibial tunnels posterior and parallel to the slope of the intercondylar roof. Reconstructions with anterior tibial tunnels resulted in graft impingement that caused increases in graft signal intensity. This increase demonstrates a clear association between surgical technique and the subsequent MR appearance of the graft.     

Pancreatic carcinoma versus chronic pancreatitis: dynamic MR imaging.

PURPOSE: To determine if dynamic gadolinium-enhanced magnetic resonance (MR) imaging can distinguish chronic pancreatitis from pancreatic carcinoma. MATERIALS AND METHODS: A retrospective review of MR and pathology examination findings was performed for 24 patients with pancreatic ductal adenocarcinoma and seven with chronic pancreatitis who underwent dynamic gadolinium-enhanced breath-hold spoiled gradient-echo imaging. Arterial, portal, and delayed phase images were obtained after injection of gadopentatate dimeglumine. The MR images of 14 patients without clinical evidence of pancreatic disease were also reviewed as controls. Signal intensity (SI) was measured on the precontrast (pre) and gadolinium-enhanced (post) images of the area of the pancreas sampled at biopsy and of the nontumorous pancreas. Percentage enhancement was defined as SIpre/SIpost x 100. RESULTS: Normal pancreas showed rapid enhancement that peaked in the arterial or portal phase. For both diseases, T1-weighted images showed hypointense masses with progressive enhancement (differences were significant [P < .05] on only delayed fat-saturated images). Differences in enhancement between either disease state and normal pancreas were significant for at least one phase. Nontumorous pancreas in patients with carcinoma showed gradual enhancement that was significantly different from that of normal pancreas. CONCLUSION: Chronic pancreatitis and pancreatic carcinoma show abnormal pancreatic enhancement, but the two were not distinguished on the basis of degree and time of enhancement.     

Pancreatic signal intensity on Tl- weighted fat saturation MR images: Clinical correlation

To determine whether signal intensity (SI) of the pancreas that was less than that of liver on Tl- weighted fat-suppressed (T1FS) magnetic resonance (MR) images could be used to help predict the presence of pancreatic disease, three blinded independent observers graded pancreatic SI relative to liver and/or renal cortex in 89 patients on T1PS images. Results were correlated with patient age and diagnosis. Among the 47 patients with a clinically normal pancreas, pancreatic SI was higher than that of liver in 42 and equal to that of liver in the rest, none of whom had evidence of fatty pancreas. These five patients had a mean age of 71 years, compared with 55 years for patients whose pancreas was more intense than liver (P=.02). Of the 42 patients with a clinically abnormal pancreas, only eight had pancreatic SI higher than that of liver. The positive predictive value for normal pancreas of an SI higher than that of liver was 84% and the positive predictive value for abnormal pancreas of an SI less than or equal to that of liver was 88%, with an overall accuracy of 86%. If normal pancreatic SI is defined as higher than that of liver for patients younger than 60 years and as equal to or higher than that of liver for patients older than 60 years, the positive predictive value of normal SI becomes 76%, the positive predictive value of decreased SI becomes 100%, and the overall accuracy becomes 83%. Pancreatic SI less than that of liver correlates highly with pancreatic disease, especially in younger patients. Mildly decreased pancreatic SI, equal to that of nondiseased liver, is common in older persons without clinically apparent pancreatic disease.     

MR imaging of small hepatocellular carcinoma: effect of intratumoral copper content on signal intensity.

Small (less than 5 cm) hepatocellular carcinomas (HCCs) in 45 of 112 patients (40.2%) had a high-signal-intensity pattern (relative to that of liver) on T1-weighted magnetic resonance (MR) images. To identify the cause of this pattern, specimens of histologically defined HCC with high- (n = 21), iso- (n = 6), and low- (n = 22) intensity patterns were histologically evaluated for paramagnetic ionic forms of metals. The incidence of steatosis, clear cell formation, and copper accumulation was statistically (P less than .01) higher in tumors with the high-intensity pattern than in those with other patterns. Of 17 HCCs that stained positive for copper, 16 (94%) had a high-intensity pattern; the pattern of one tumor (6%) was isointense. All 21 tumors with the high-intensity pattern had at least one of the findings of steatosis, clear cell formation, or copper accumulation. These features were evident in only four of 28 tumors (14%) with an iso- or low-intensity pattern. Thus, copper accumulation might be a cause of the high-intensity pattern on T1-weighted images of small HCCs.     

1.5-T MR imaging of pituitary microadenomas: technical considerations and CT correlation

Thirty-seven patients with suspected pituitary tumors were evaluated prospectively with MR imaging at 1.5 T. MR detected a microadenoma at its correct location in all eight patients who underwent transsphenoidal surgery, while CT showed a focal abnormality in the correct location in only four of the eight patients. In patients who were clinically and endocrinologically considered to harbor a microadenoma, MR detected a focal pituitary signal abnormality in 83% and CT demonstrated a focal density abnormality in 42%. Infundibular displacement, focal gland convexity, and sellar-floor abnormality were seen equally well with CT and MR. MR imaging protocol included sagittal T1-weighted spin-echo, coronal inversion-recovery, and coronal spin-echo or cardiac-gated spin-echo images. Although inversion-recovery images were superior in detecting focal pituitary lesions, some microadenomas were better seen on T2-weighted images. Cardiac-gated spin-echo images showed focal pituitary lesions better than ungated images did. Our technique demonstrates MR's superior sensitivity to CT in detecting a pituitary microadenoma.     

MR imaging of hepatocellular carcinoma: relationship between lesion size and imaging findings, including signal intensity and dynamic enhancement patterns

PURPOSE: To assess the relationship between lesion size and MR imaging findings of pathologically-proven hepatocellular carcinoma (HCC). MATERIALS AND METHODS: In a retrospective, single-center study, 37 consecutive patients were identified between 1999 and 2005 that underwent preoperative MRI and surgical resection of HCC. A total of 47 lesions (mean size = 6.85 cm, range = 1-25 cm) were assessed for signal intensity (SI), enhancement patterns, and secondary morphologic features. Interobserver rating, percentage enhancement, and contrast-to-noise-ratio (CNR) were determined. Lesions were assessed for combinations of typical MRI features. Regression analysis was used to assess relations between MRI findings and tumor size. RESULTS: On fat-suppressed T2-weighted (T2w) fast-spin-echo, smaller lesions had lower SI compared to larger lesions (P < 0.05). In the arterial phase, smaller lesions showed significantly higher percentage enhancement compared to larger lesions (P < 0.05). In the delayed phase, smaller lesions showed less pronounced washout (P < 0.05). Heterogeneity of the lesions, including fatty infiltration, internal nodules, or mosaic pattern, was observed significantly more frequently in larger lesions (P < 0.001). The classic combination of high T2w signal, strong arterial enhancement, and delayed phase washout was present in 23 of 44 lesions (52%). CONCLUSION: Smaller HCC often showed lower SI on T2w, more intense arterial enhancement, and less pronounced delayed washout compared to larger HCC.     

MR enteroclysis: technical considerations and clinical applications

Magnetic resonance enteroclysis (MRE) is an emerging technique for the evaluation of small bowel abnormalities. Adequate luminal distention, achieved by the administration of iso-osmotic water solution through a nasojejunal catheter, in combination with ultrafast sequences, such as single-shot turbo spin echo, true fast imaging with steady precession, half-Fourier acquired single-shot turbo spin echo, and 3D fast low-angle shot, results in excellent anatomic demonstration of the small bowel. Magnetic resonance fluoroscopy can be performed during MRE examination and might be useful in studying low-grade stenosis or motility-related disorders. Magnetic resonance enteroclysis is very promising in detecting the number and extent of involved small bowel segments in patients with Crohn’s disease, and in disclosing lumen narrowing and extramural manifestations and complications of the disease. Initial experience shows that MRE is very efficient in the diagnosis of small bowel tumors and can be used in the evaluation of small bowel obstruction. Electronic Publication     

Cirrhotic nodules: association between MR imaging signal intensity and intranodular blood supply

PURPOSE: To retrospectively determine whether there is a relationship between the intranodular blood supply evaluated at computed tomography (CT) during arterial portography (CTAP) and CT during hepatic arteriography (CTHA) and the magnetic resonance (MR) imaging signal intensity of nodules associated with cirrhosis. MATERIALS AND METHODS: Neither institutional review board approval nor informed consent was required for retrospective reviews of medical records and images. One hundred fourteen hepatocellular nodules 10 mm or greater in largest diameter in 58 patients (39 men, 19 women; mean age, 61 years) with cirrhosis were evaluated at CTAP, CTHA, and MR imaging. The CTAP and CTHA nodule findings were divided into three main types: Type A nodules were isoattenuating at CTAP and hypoattenuating at CTHA; type B nodules, slightly hypoattenuating at CTAP and hypoattenuating at CTHA; and type C nodules, strongly hypoattenuating at CTAP and hyperattenuating at CTHA. The relationships between the CTAP and CTHA findings and the MR imaging signal intensity among these nodules were analyzed by using the chi(2) test. RESULTS: On T1-weighted MR images, 27 (63%) of 43 type A nodules were hyperintense, nine (39%) of 23 type B nodules were isointense, and 19 (48%) of 40 type C nodules were hypointense; differences were not significant. On T2-weighted MR images, 31 (72%) of 43 type A nodules were hypointense (P < .05), 12 (52%) of 23 type B nodules were isointense, and 34 (85%) of 40 type C nodules were hyperintense (P < .05). CONCLUSION: There was a significant association between intranodular blood supply and nodule signal intensity on T2-weighted MR images. However, study findings did not show whether the blood itself (ie, blood volume or blood flow amount) directly influenced the signal intensity.     

High signal intensity lesions of the chest in MR imaging

The majority of pathologic lesions in the lung and mediastinum have relatively long T1 and T2 relaxation times and consequently yield medium to low signal intensity on T1-weighted images. Pulmonary lesions with high signal intensity on T1-weighted images are unusual and raise a special group of diagnostic considerations. In the current study, a mass with a lesion/fat signal intensity ratio of greater than 0.7 on a T1-weighted sequence was considered high signal intensity. The nature of these masses was ganglioneuroma or ganglioneuroblastoma (n = 3), atrial lipoma (lipomatous atrophy of the interatrial septum) (n = 3), pheochromocytoma (n = 2), bronchogenic cyst (n = 2), lymphangioma (n = 1), teratoma (n = 1), and a variety of primary and metastatic tumors of the mediastinum and lung. A single pathologic structure of these lesions was not present, but rather several underlying tissue compositions were noted, including fat, subacute hemorrhage, myxoid material, and cellular composition with high cytoplasmic/nuclear ratio. Thus, high signal intensity lesions of the thorax on T1-weighted images should suggest a number of differential diagnoses.     

Monitoring of the neonate undergoing MR imaging: technical considerations. Work in progress

Instrument monitoring of vital signs in neonates undergoing magnetic resonance (MR) imaging can be difficult because of the unique environmental restrictions imposed by the imager. The authors present their experience with monitoring more than 50 newborn infants and discuss the interaction of monitoring devices with the MR imager. Several MR-compatible monitors allow continuous evaluation of body temperature, heart rate, blood pressure, and auscultation of heart sounds and respiration in mechanically ventilated infants. Signal-to-noise (S/N) ratio measurements taken during imaging of the head of an infant with these monitors in place did not differ appreciably from the ratio obtained during imaging without monitors. Tip angles should be optimized to account for widely varying head size among neonates, since adverse monitoring effects are significantly compounded by improper tip angle adjustment.     

RARE imaging: a fast imaging method for clinical MR

Based on the principles of echo imaging, we present a method to acquire sufficient data for a 256 X 256 image in from 2 to 40 s. The image contrast is dominated by the transverse relaxation time T2. Sampling all projections for 2D FT image reconstruction in one (or a few) echo trains leads to image artifacts due to the different T2 weighting of the echo. These artifacts cannot be described by a simple smearing out of the image in the phase direction. Proper distribution of the phase-encoding steps on the echoes can be used to minimize artifacts and even lead to resolution enhancement. In spite of the short data acquisition times, the signal amplitudes of structures with long T2 are nearly the same as those in a conventional 2D FT experiment. Our method, therefore, is an ideal screening technique for lesions with long T2.     

Temporomandibular joint: morphology and signal intensity characteristics of the disk at MR imaging

Magnetic resonance (MR) imaging has been used in the temporomandibular joint (TMJ) primarily to define the disk position. This report examines altered morphology and signal intensity characteristics of the TMJ disk as they relate to the severity of internal derangement. Two hundred sixteen joints in 133 patients with a history of such derangement. were imaged with MR. Disk position, signal intensity, morphology, and the presence of osteoarthritis were determined for each joint. The normal disk was not anteriorly displaced and had a normal "bow-tie" shape. A grade 1 disk was anteriorly displaced and had a normal shape; a grade 2 disk was anteriorly displaced and had an abnormal shape. Forty (19%) joints were considered normal; none of these exhibited osteoarthritis. One hundred thirty-nine (64%) joints were grade 1; osteoarthritis was found in 17%. Thirty-seven (17%) were grade 2; osteoarthritis was found in 95%. All forty normal joints had high or intermediate signal intensity in the disk. Osteoarthritic joints had a higher percentage of disks with diminished intensity (P less than .0001). Severe or untreated osteoarthritis is known to be a complication of TMJ internal derangements; hence this grading system seems to correlate with the severity of internal derangement.     

Articular cartilage: correlation of histologic zones with signal intensity at MR imaging

Zones of high and low signal intensity on magnetic resonance (MR) images of articular cartilage were correlated with the four histologic zones normally found in such cartilage. Grossly normal articular cartilage from knees and ankles of a fresh cadaver were used in the study. The three zones identified on MR images included a low-intensity zone near the articular surface, a zone of higher signal intensity next to that, and a second zone of low intensity that was deep to the two others. The location of the superficial low-intensity zone corresponded to dense, tangentially oriented layers of collagen in the superficial histologic zone. Higher signal intensity deep to the superficial low-intensity zone correlated with cartilage in the transitional zone. The deep low-intensity zone correlated with a combination of deep radiate and calcified cartilage and cortical bone. Results of this study indicate that, with high resolution, MR imaging may demonstrate three zones of differing signal intensity in articular cartilage. The superficial low-intensity zone may be a useful marker of the surface of normal articular cartilage.