NMR imaging of leg tumors

Pulse sequences as tissue property filters (TP-filters): a way of understanding the signal, contrast and weighting of magnetic resonance images

This paper describes a quantitative approach to understanding the signal, contrast and weighting of magnetic resonance (MR) images. It uses the concept of pulse sequences as tissue property (TP) filters and models the signal, contrast and weighting of sequences using either a single TP-filter (univariate model) or several TP-filters (the multivariate model). For the spin echo (SE) sequence using the Bloch equations, voxel signal intensity is plotted against the logarithm of the value of the TPs contributing to the sequence signal to produce three TP-filters, an exponential ρm-filter, a low pass T1-filter and a high pass T2-filter. Using the univariate model which considers signal changes in only one of ρm, T1, or T2 at a time, the first partial derivative of signal with respect to the natural logarithm of ρm, T1 or T2 is the sequence weighting for each filter (for small changes in each TP). Absolute contrast is then the sequence weighting multiplied by the fractional change in TP for each filter. For large changes in TPs, the same approach is followed, but using the mean slope of the filter as the sequence weighting. These approaches can also be used for fractional contrast. The univariate TP-filter model provides a mathematical framework for converting conventional qualitative univariate weighting as used in everyday clinical practice into quantitative univariate weighting. Using the multivariate model which considers several TP-filters together, the relative contributions of each TP to overall sequence and image weighting are expressed as sequence and imaging weighting ratios respectively. This is not possible with conventional qualitative weighting which is univariate. The same approaches are used for inversion recovery (IR), pulsed gradient SE, spoiled gradient echo (SGE), balanced steady state free precession, ultrashort echo time and other pulse sequences. Other TPs such as susceptibility, chemical shift and flow can be included with phase along the Y axis of the TP-filter. Contrast agent effects are also included. In the text TP-filters are distinguished from k-space filters, signal filters (S-filters) which are used in imaging processing as well as to describe windowing the signal width and level of images, and spatial filters. The TP-filters approach resolves many of the ambiguities and inconsistencies associated with conventional qualitative weighting and provides a variety of new insights into the signal, contrast and weighting of MR images which are not apparent using qualitative weighting. The TP-filter approach relates the preparation component of pulse sequences to voxel signal, and contrast between two voxels. This is complementary to k-space which relates the acquisition component of pulse sequences to the spatial properties of MR images and their global contrast.

Magnetic Resonance Imaging in Diffuse Malignant Bone Marrow Diseases

Twenty-four patients with malignant bone marrow involvement or polycythemia vera, 8 patients with reactive bone marrow and 7 healthy individuals were examined with spin-echo magnetic resonance imaging at 0.35 T and 0.5 T. Signs of an increased longitudinal relaxation time, T1, were found when normal bone marrow was replaced by malignant cells, polycythemia vera or reactive marrow. A shortened T1 was indicated in 4 patients in bone marrow regions treated by radiation therapy; the marrow was most likely hypocellular in these cases. The estimated T1 relaxation times were highly correlated to the cellularity of the bone marrow as assessed by histology. Among patients with close to 100 per cent cellularity neither T1 nor T2 discriminated between the various malignancies or between malignant and reactive, non-malignant bone marrow. Characterization of tissues in terms of normalized image intensities was also attempted, the motive being to avoid approximations and uncertainties in the assessment of T1 and T2. The normalization was carried out with respect to the image of highest intensity, i.e. the proton density weighted image. The results were in agreement with those for T1 and T2. It was concluded that MRI is valuable for assesssing bone marrow cellularity, but not for differentiating between various bone marrow disorders having a similar degree of cellularity.

The use of MRI in the diagnosis of benign and malignant bone and soft tissue tumours

Two hundred and thirty four suspected primary bone and soft tissue tumours were investigated using plain films and MRI. The MR appearance of 200 of these tumours was assessed with respect to the intensity of the lesion, the homogeneity of the tumour, the presence or absence of a capsule or lobulation, whether the tumour was whorled or not and whether it contained either fluid or blood. Apart from benign lipomas and some malignant myxoid liposarcomas, however, it seemed virtually impossible to tell one tumour from another and in many cases to differentiate a benign lesion from a malignant tumour using MRI alone. We recommend that the workup of a suspected soft tissue tumour should be initially by MR scanning and that the workup of a suspected malignant bone tumour should be plain films followed by an MRI scan.

Destructive changes of the spine in magnetic resonance imaging

Twenty-nine patients were examined by magnetic resonance imaging for various lesions of the spine. The results of these scans were compared with those of plain radiographs, computertomographs, and radionuclide bone scans. The findings were substantiated by intraoperative or histological findings in 18 cases. The MRI scans proved to be very sensitive in the detection of a wide spectrum of morphological changes of the bone marrow of vertebral bodies. Characteristic changes of the signal patterns for inflammatory and tumorous lesions were not observed. The differentiation of these lesions will presently continue to have to be based on morphological criteria.

Magnetic resonance imaging: comparison of four pulse sequences in assessing primary bone tumours

A prospective magnetic resonance imaging (MRI) study was carried out in 13 patients (19 examinations) with primary bone tumours to assess the relative value of each of four pulse sequences in showing the extent and nature of the lesion. The four pulse sequences used were a T1-weighted spin-echo (SE544/44), a T2-weighted spin echo (SE1500/80), a short TI inversion recovery (STIR) (IR500/100/44), and a partial saturation (PS) (PS500/22) with field echo data collection. For soft tissue disease the combination of PS and STIR gave better definition of the boundary of the tumour than the more conventional T1 and T2-weighted spin echo sequences. For the demonstration of bone cortex, periosteal change and calcification, T1 and T2-weighted spin echo sequences were better. However, for calcified tissues, plain radiographs were better than either MRI combination. On the assumption that plain films will be available in all cases, PS and STIR sequences could therefore be substituted for T1 and T2-weighted spin echo sequences allowing an increase in soft tissue detectability for lesions in both red and yellow marrow.

The role of magnetic resonance imaging in the evaluation of Ewing sarcoma. A report of 27 cases

The experience with magnetic resonance imaging (MRI) in 27 patients with Ewing sarcoma is reported and compared with computed tomography (CT) and plain films. Plain radiography proved to be the best imaging method to asses probable histological diagnosis in all cases (n = 6). For the evaluation of chemotherapeutic response (n = 4), CT and MRI gave the same information about the variation in size of the tumor. In this small series, the high signal in T2 weighted images was not altered significantly by therapy. In preoperative evaluation (n = 14), MRI gave better information than CT of soft tissue involvement and extension within the bone marrow in two cases each. The ability of MRI to accurately define extension through the epiphyseal plate in two cases permitted limb salvage which otherwise would not have been possible. In the long-term follow-up (n = 12), three patients without recurrence one year after therapy showed a low signal in the surgical area in T2 weighted images. Nine patients had a high signal in T2 weighted images: four were reactive lesions, two had obvious recurrence, and one was a hematoma. In the two remaining cases plain films and CT were normal, in the presence of both active tumor and reactive lesions. It was not possible with MRI to differentiate active tumor from reactive change, even after Gd-DTPA infusion.

Prenatal diagnosis and management of bilateral hydronephrosis

This report reviews the management of the fetus with congenital hydronephrosis (CH), a challenging diagnostic and therapeutic problem. Experimental models of obstructive uropathy have produced histologic changes similar to those seen in kidneys of human neonates with congenital hydronephrosis. Relief of obstruction in utero in these models has been shown to prevent some of the dysplastic changes caused by obstruction. These studies have formed the theoretical basis for in utero decompression to restore amniotic fluid dynamics to prevent death from pulmonary hypoplasia, and reverse or arrest dysplastic morphogenesis. The development of prognostic criteria has greatly aided in selection of appropriate fetuses for intervention. These criteria include: (1) Na less than 100 mEq/l; (2) Cl less than 90 mEq/l; (3) osmolarity less than 210 mosmol; (4) sonographic appearance of the fetal kidneys; (5) amniotic fluid status; (6) urine output at fetal bladder catheterization. All fetuses should have ultrasonography to exclude other anomalies, and karyotype analysis to exclude chromosomal abnormality. If amniotic fluid volume is normal, the pregnancy is followed with serial ultrasound examinations. If oligohydramnios develops, a prognostic evaluation is performed, including fetal bladder catheterization. If the fetus has poor residual renal function, on the basis of prognostic criteria, appropriate counseling may be given. If the fetus has good residual renal function, depending on lung maturity, it can be delivered early for corrective surgery. If diagnosed prior to lung maturity in utero, decompression by either vesicoamniotic shunting or open fetal surgery may be attempted in the highly selected case.(ABSTRACT TRUNCATED AT 250 WORDS)

A review of applications of MRI in soft tissue and bone tumors

Due to excellent soft tissue contrast and multiplanar imaging capability, MRI is assuming a major role in recognition, staging, and treatment planning of soft tissue and bone tumors. Direct sagittal, coronal, and axial images permit assessment of intraosseous and extraosseous extension of tumors and their relationship to the joints and neurovascular structures, and detection of "skip" lesions. MRI allows improved detection of recurrent tumors in the presence of non-ferromagnetic metallic implants as compared to CT. In the evaluation of soft tissue tumors, MRI is more sensitive than CT and allows differentiation among fat, muscle, tendon, bone, and vascular structures based on signal characteristics. Over a period of 18 months, 100 soft tissue masses and bone tumors were evaluated using MRI. Spin echo sequences with T1 and T2 weighted images were most valuable in differentiating normal and abnormal tissues. Calculated comparative measurements of relaxation times showed no reliable difference between benign and malignant tumors.

The role of magnetic resonance imaging in giant cell tumor of bone

In six cases of giant cell tumor the magnetic resonance (MR) images obtained with various pulse sequences and field strengths were compared to the corresponding computed tomography (CT) scans and plain roentgenograms. MRI was superior to CT and plain films in demonstrating areas of tissue inhomogeneity within the tumor as well as soft tissue extension. CT was superior in demonstrating cortical thinning. Multiplanar imaging capability and visualization of articular cartilage may demonstrate intra-articular tumor spread. The characteristic MRI findings with short TR/TE (T1-weighting) and long TR/TE (T2-weighting) are described. We also describe one case where serial MR scans were used to assess response to therapy.

Magnetic resonance imaging of knee hyaline cartilage and intraarticular pathology

Injuries to the hyaline cartilage of the knee joint are difficult to diagnose without invasive techniques. Even though these defects may be the most important prognostic factors in assessing knee joint injury, they are usually not diagnosed until arthrotomy or arthroscopy. Once injuries to hyaline cartilage are found and/or treated, no technique exists to follow these over time. Plain radiographs, arthrograms, and even computed tomography fail to detail most hyaline cartilage defects. We used magnetic resonance imaging (MRI) to evaluate five fresh frozen cadaver limbs and 10 patients whose pathology was known from arthrotomy or arthroscopic examination. Using a 0.35 Tesla superconducting magnet and spin-echo imaging technique with a head coil, we found that intraarticular fluid or air helped to delineate hyaline cartilage pathology. The multiplane capability of MRI proved to be excellent in detailing small (3 mm or more) defects on the femoral condyles and patellar surface. Cruciate ligaments were best visualized on sagittal oblique projections while meniscal pathology was best seen on true sagittal and coronal projections. MRI shows great promise in providing a noninvasive technique of evaluating hyaline cartilage defects, their response to treatment, and detailed anatomical information about cruciate ligaments and menisci.

Application of magnetic resonance imaging and computerized tomography as an adjunct to the surgical management of soft tissue sarcomas

Magnetic resonance imaging (MRI) and computed tomography (CT) scans of 53 evaluable patients with biopsy-proven soft tissue sarcomas were reviewed and compared with operative results to ascertain the accuracy for each imaging modality to predict resectability. Location of soft part sarcomas included: abdomino-pelvic (3), retroperitoneal (7), extremity (35), and other anatomic sites (8). MRI was observed to have greater accuracy than CT to preoperatively predict resectability (96.2% vs. 75.5%, respectively, p = 0.0034) following three-dimensional, multiplanar evaluation. Further, MRI was judged to have superior sensitivity to CT (95.6% vs. 73.3%, respectively, p = 0.006) and equivalent specificity (100% vs. 87.5%, respectively, p = 0.125). MRI represents a sophisticated diagnostic imaging technique to differentiate normal tissue from soft tissue sarcomas with superior contrast resolution in multiplanar imaging. MRI is considered to be the imaging modality of choice for these tumors with the advantage of not exposing the patient to ionizing irradiation or intravenous contrast agents to delineate contiguous structures.

Magnetic resonance imaging versus computed tomography in the evaluation of soft tissue tumors of the extremities

Twenty patients with extremity soft tissue tumors were prospectively evaluated with magnetic resonance imaging (MRI) and computed tomography (CT) scans with subsequent anatomic correlation of surgical findings. MRI and CT had a similar percentage of accuracy in assessing tumor relationship with major neurovascular (80% and 70%, respectively) and skeletal (80% and 75%, respectively) structures. MRI was significantly better than CT in displaying contrast between tumor and muscle when using the T2 weighted spin echo (SE) (p2 less than 0.002) and inversion recovery (IR) (p2 less than 0.005) pulse sequences. MRI and CT were comparable in demonstrating contrast between tumor and fat. The contrast between tumor and vessel was better displayed by MRI compared with CT when using the T1 weighted SE (p2 less than 0.001) and T2 weighted SE (p2 less than 0.001) pulse sequences. T1 and T2 values were measured on fresh tumor and normal tissue samples and were used to predict relative contrast on different MRI pulse sequences using isosignal contour plots. MRI appears to offer several advantages over CT in the evaluation of extremity soft tissue tumors.

Fluid-fluid levels in giant cell tumors of bone: report of two cases

Fluid-fluid levels have been described in association with aneurysmal bone cysts, telangiectatic osteosarcoma, and a chondroblastoma. We report two cases of giant cell tumors of bone with fluid-fluid levels identified by computed tomography and, in one case, by magnetic resonance imaging. This finding has not previously been associated with giant cell tumors. The radiographic features of the fluid-fluid levels cannot be distinguished from those reported in other osseous neoplasms.

Magnetic resonance imaging of primary tumours and tumour-like lesions of bone

The experience with magnetic resonance imaging (MRI) of 81 patients with primary bone tumours and tumour-like lesions is reported. MRI proved to be a sensitive method of detecting primary bone tumours. Intramedullary and extraosseous parts of bone tumours were delineated better than by plain films and computed tomography (CT). Surgical clips and Harrington rods did not appreciably limit the estimation of tumour recurrence. MRI provided definite advantages compared to CT in the surgical staging of bone tumours and tumour-like lesions. MRI was found to be an imaging method with low specificity. Differentiation of tissue components, such as haematoma, fat, necrosis, and cystic areas, led to a specific diagnosis only in rare cases. Plain films and CT were found to be superior to MRI in assessing the biological activity and the differential diagnosis of bone tumours and tumour-like lesions.

Magnetic resonance imaging of the extremities

MRI offers the best soft-tissue contrast of any radiographic modality. It provides better delineation of soft-tissue tumors and bone tumors than CT, but several different pulse sequences may be needed to make optimum use of this potential. MRI is also well suited to evaluation of abnormalities of the bone marrow, as it is not subject to the dense bone artifact that may occur with CT. Changes in signal intensity may be used to gain some appreciation of blood flow in the major vessels. Examples of each of these applications are discussed.

Present and future clinical position of magnetic resonance imaging

Following the appearance of the first clinical reports on magnetic resonance imaging (MRI) five years ago, stock is now taken of the clinical value of this new imaging technique and its future potential is assessed. Contrast and spatial resolution as well as specificity of CT and MRI are compared and the multiplicity of sequences, multiplanar facility, and flexibility of MRI over CT are analyzed. The dominant role of MRI over CT in the study of some disorders of the central nervous system is assessed and so are the limitations. In this context the usefulness of paramagnetic contrast media is also evaluated. The difficulties of body imaging are discussed. The solutions of overcoming cardiac and respiratory motions artifacts are evaluated and areas where MRI has already achieved a high level of clinical application are identified. Mention is made of additional information obtainable by MRI as, for example, sodium imaging, blood flow measurements, and the combination of imaging with spectroscopy. Possible future prospects are analyzed, for example, high-resolution imaging, alternative sequences to those commonly used in particular the short TI IR sequence (STIR), chemical shift imaging, and the possibility of utilizing susceptibility for imaging purposes. The problem of radiofrequency receiver coil design is approached and so is the choice of magnetic fields. The future relationship of MRI with established noninvasive techniques is stressed.

Magnetic resonance imaging of extremity masses

Twenty-seven adults with extremity masses were examined by magnetic resonance imaging (MRI). In 26/27 cases, computed tomography (CT) scans were available for comparison. Imaging with multiple pulse sequences is necessary to optimize diagnosis by MRI. Advantages of MRI include direct sagittal and coronal imaging, demonstration of vessels without contrast and superior soft tissue contrast. CT better demonstrates cortical destruction and small calcifications, which may be important in some cases. For noncalcified masses without bone involvement, MRI is equivalent or superior to CT. In those cases with calcification or bone involvement, MRI plays a role complementary to CT.

Magnetic resonance imaging of lesions of synovial origin

Three patients with histologically differing lesions of synovial origin and two with synovial cysts, one of which was a dissecting popliteal cyst, were examined by magnetic resonance imaging (MR) and computerized tomography (CT). The three histologically proven synovial lesions were synovial sarcoma, diffuse giant cell tumor of tendon sheath, and synovial chondromatosis. In two of the five patients MR provided better anatomic and morphologic appreciation than CT, while in the others they were of equal value. CT demonstrated calcification in two of the lesions while on MR calcification could be identified in only one patient where it outlined the mass. MR did not demonstrate calcification in the substance of the diffuse giant cell tumor of tendon sheath. Coronal, transverse, and sagittal images of magnetic resonance graphically demonstrated the extent of the soft tissue masses and their relationship to bone, vessels, and soft tissue structures. Synovial sarcoma had a shorter T1 than diffuse giant cell tumor of tendon sheath (these two lesions being of comparable size) and also had a uniformly longer T2. The dissecting popliteal cyst showed the most intense signals on the T1 weighted images, while the uncomplicated synovial cyst showed a long T1. On the T2 weighted images, each type of cyst showed a long T2. The variance and overlap of intensity of MR signals suggest limited specificity in predicting the histologic nature of the synovial lesion.

Magnetic resonance imaging in pediatrics

Magnetic resonance imaging has potentially broad applications in pediatric practice. Although further studies are needed to determine its exact role in comparison with the other imaging modalities, magnetic resonance has shown increased sensitivity in lesion detection in many disease processes. Since MR does not use ionizing radiation and does not require intravenous contrast to identify vascular structures, it becomes an ever more attractive imaging tool for pediatric diagnosis. Thus, the early results of MR imaging have shown promise and the future of MR appears exciting.

Magnetic resonance imaging of the musculoskeletal system

During the last few years, the interest and experience in magnetic resonance imaging (MRI) of the musculoskeletal system has increased rapidly. Modern equipment with good spatial resolution and multislice, multiecho technique provides detailed information on the joints, soft tissue and spine within a reasonable examination time, both with superconductive and resistive systems. Today, MRI is an established technique for examination of musculoskeletal tumors and aseptic bone necrosis and it has proven to be of great value for evaluation of diseases of the spine and spinal canal. The definite place of MRI within musculoskeletal diagnostic imaging is not yet settled, but its potential is great, and it will have an important role in the future.

Magnetic resonance imaging of the extremities. I. Technique for depiction of normal anatomy

In 3 adult volunteers, the appropriate technique for depicting the normal anatomy with magnetic resonance imaging was tested. At 0.15 T, the images obtained in different anatomic planes with different pulse sequences and radiofrequency coils were compared. Spin echo technique with a repetition time of 500 ms and an echo time of 30 ms appeared as a good compromise, giving excellent depiction of both soft tissues and joints, especially if a surface coil was used. The optimum anatomic plane varied with the structures examined.

Magnetic resonance imaging of fluid levels in an aneurysmal bone cyst and in anticoagulated human blood

Magnetic resonance imaging (MRI) demonstrated a fluid level within an aneurysmal bone cyst (ABC). Since the ABC contained gross blood at operation, an anticoagulated human blood sample was studied by MRI also, and a fluid level was again clearly visible. MRI pulse sequences emphasizing T1 contrast showed the fluid levels most clearly in both the ABC and the blood. Sequences emphasizing T2 contrast showed homogeneous, bright signals in the ABC and in the blood, with no visible fluid level in the ABC and a nearly invisible one in the blood. In the blood sample, the calculated plasma T1 value was 1585 ms, and that of the red cells was 794 ms.

Magnetic resonance imaging in a patient with soft tissue sarcoma of the leg

A patient with a high grade malignant fibrous histiocytoma of the leg is presented. Staging studies included a 99mTc diphosphonate bone scan, an arteriogram, a computed tomogram (CT) and a proton (1H) magnetic resonance imaging (MRI) study. By manipulating imaging parameters to enhance contrast between normal and neoplastic tissues, the latter technique more accurately delineated the extent of the soft tissue sarcoma than the other imaging modalities. Although spatial resolution of MRI is inferior to that of CT, it is likely that MRI will provide important information relative to surgical staging and preoperative planning that is not available from current imaging techniques.

Magnetic resonance imaging of bone and soft tissue tumors: early experience in 31 patients compared with computed tomography

In 31 patients with 21 soft tissue and 10 bone tumors, magnetic resonance imaging (MRI) and computed tomography (CT) were equally effective in delineating the margins of most soft tissue tumors, and the margins of bone tumors from fat and adjacent normal bone. However, MRI was superior to CT in delineating bone tumors from adjacent muscle, and in showing the relationships to bone of the deep margins of some soft tissue tumors. This was true because the quality of CT images around thick cortical bone often was severely degraded by streak artifact, which does not occur in MRI. Excellent anatomic detail was achieved on MRI by spin echo pulse sequences with short repetition times. Bone tumors were delineated best by spin echo 1000/30 images, and soft tissue tumors by spin echo 1000/30 or inversion recovery images.

Magnetic resonance imaging: application in musculoskeletal infection

Forty-two patients with clinically suspected osteomyelitis were examined using magnetic resonance imaging (MRI). Twenty-seven patients (64%) had previous surgery or fracture, and 15 (36%) were referred for differentiation of acute osteomyelitis from bone tumors or other pathologic conditions. MRI was compared with computed tomography in 12 cases and with 111In-labeled leukocytes scans in 22. With MRI, 92% of proved infections were detected, and bone and soft-tissue changes were more evident than with routine radiographs, tomography, or computed tomography. In patients with negative cultures and no previous surgery or fracture, it was difficult for MRI to differentiate operative changes from infection. In these patients, 111In-labeled leukocyte images were more specific than MRI.

Limitations of computed tomography following excisional biopsy of soft tissue sarcomas

Twenty-one patients were evaluated by computed tomography (CT) following complete or incomplete excisional biopsy of soft tissue sarcomas. Since the surgical margins were inadequate, additional treatment was required, and CT was intended to identify and delineate any residual tumor. Thirteen patients had no palpable mass in the operative area. In these, eleven CTs showed no tumor, but microscopic tumor was found in seven. The other two CTs showed masses, but both proved to be hematomas. Eight patients had palpable masses and seven proved to be residual tumor. Of these, two CTs failed to show the residual tumor. Five CTs correctly identified residual tumor, but two of them failed to show the entire tumor extent. Computed tomography did not correctly predict the presence or absence of microscopic residual tumor when there was no palpable mass, and was at least partly inaccurate in delineating residual tumor in four of eight patients with palpable masses.

Magnetic resonance. Principles and applications

Magnetic resonance is a unique, noninvasive imaging modality which allows direct, multiplanar imaging and the possibility of obtaining biochemical information in vivo. Presently, MR appears most applicable to the evaluation of central nervous system abnormalities. The high sensitivity of MR in the evaluation of intracranial pathology suggests that it may eventually replace CT for many suspected diseases, if future investigations are able to improve its specificity. As previously noted, MR may be more diagnostic than other radiologic studies in the evaluation of suspected Chiari malformation, syringomyelia, congenital abnormalities, tumors of the spinal axis, and disc space infection. In the chest, MR appears to be more accurate than CT in the determination of the extent of mediastinal tumor, but at present cannot replace CT because of the lack of experience in imaging parenchymal nodules and benign diseases. MR of the breast is promising, but the size of the lesion may prove to be a limiting factor with magnetic field strengths commonly being utilized. There are inherent difficulties in the evaluation of cardiac disease with MR, but it offers a noninvasive method of investigating congenital heart disease and may provide valuable information in suspected myocardial ischemia and altered cardiac function. MR provides a new method of evaluating the vascular system, both in terms of providing anatomic information on large and medium-sized vessels and flow analysis. In the abdomen, MR appears to be most sensitive in the evaluation of suspected hepatic masses, but as with the brain, greater specificity will be needed to replace CT. At the present time, MR offers no distinct advantage over conventional imaging modalities in the evaluation of pancreatic disease, it maybe more accurate than CT in the staging of renal cell carcinoma. Larger studies are needed to determine the role of MR in the investigation of retroperitoneal adenopathy and adrenal abnormalities. In the pelvis, MR offers the hope of earlier diagnosis of prostatic carcinoma and may replace CT for staging of prostatic carcinoma and transitional cell carcinoma of the bladder. Limited MR experience with benign disease of the female pelvis suggests that it is currently more accurately evaluated with ultrasound. MR appears to be highly sensitive and specific for the diagnosis of avascular necrosis and may provide an early clue in suspected osteomyelitis. Finally, in vivo MR spectroscopy may provide unique metabolic information that was unobtainable prior to the advent of magnetic resonance, if this proves to be technically feasible.