Faculty Bibliography

Subacute intracranial hematomas have variable appearances on high-field MR images. They are hyperintense on T1-weighted images owing to methemoglobin, but have variable intensities on T2-weighted images. Observation of the different high-field spin-echo MR intensity patterns of five subacute hematomas suggests that further subcategorization into different methemoglobin states may be possible. In particular, undiluted intracellular methemoglobin is hyperintense on T1-weighted images and markedly hypointense on T2-weighted images, undiluted free methemoglobin should be hyperintense on T1-weighted images and isointense or slightly hypointense on T2-weighted images, and dilute free methemoglobin is hyperintense on both T1- and T2-weighted images. However, it appears that certain regions of subacute hematomas may be difficult to differentiate, by intensity patterns alone, from melanotic melanomas or fat. We believe that, despite some limitations, MR is useful in dividing subacute intracranial hematomas into their respective methemoglobin states, and also that further subcategorization is possible

The relationship between tumor mass and vascular involvement as seen on MR imaging was examined in 11 patients with masses in the parasellar region, and the findings were correlated with CT and angiography. In six cases, MR was superior to CT and angiography in depicting the relationship of the tumor to adjacent blood vessels. In these cases, MR demonstrated tumor surrounding the blood vessel without changing the diameter of its lumen. Angiography did not reveal encasement in these cases. In four cases, both MR and angiography showed signs of vascular encasement with narrowing of the vessel's lumen. In two cases, MR was equivocal while angiography revealed vascular encasement in one case and was negative for encasement in the other. CT was less sensitive than MR in defining vascular encasement since there is usually little contrast between an enhancing tumor and the major blood vessels. Coronal scanning appeared to be the best plane of imaging and correlated well with the anteroposterior angiogram. We propose that MR is the method of choice for evaluating arterial encasement by tumors and may obviate the need for angiography in those cases in which MR is positive for a basal lesion

A retrospective study of the magnetic resonance (MR) images of the lumbar spines of 13 healthy subjects and 30 patients with degenerative changes was done. In the healthy subjects, the vertebral facets, thickness of the cartilage and ligamentum flavum, signal characteristics of the bone marrow, and size of the spinal canal were studied. In the patients with degenerative changes in one of these structures, MR images in the sagittal plane were useful in demonstrating hypertrophy of the ligamentum flavum or the vertebral facets, in grading the degree of foraminal stenosis, and in measuring the sagittal diameter of the spinal cord. MR images in the axial plane facilitated detailed analysis of the facet joint and more accurate measurements of the thickness of the ligamentum flavum and spinal canal diameter. MR images were compared with computed tomography scans in 12 patients

Acute experimental allergic encephalomyelitis, an animal model of CNS inflammatory and demyelinating disease, was produced in guinea pigs and imaged with MR. Correlation of histopathology with MR revealed acute mononuclear perivascular inflammatory changes and edema corresponding to the high-intensity abnormalities observed on long repetition time (TR) images. Our results suggest that in acute multiple sclerosis the high intensity noted on long TR images may be secondary to associated inflammatory changes rather than to demyelination

This study attempts to determine the magnitude of change in the concentration of a nonparamagnetic protein (human serum albumin) required to effect a detectable change in signal intensity on a clinical imaging unit. For a range of protein concentrations from 0-6100 mg/dl the concentration could not be predicted by inspecting the images. Measurement of displayed signal intensity failed to distinguish concentrations of 0.09-3700 mg/dl, while 6100 mg/dl gave slightly higher intensity signals. Although this low sensitivity represents expected behavior for low concentrations, the failure to differentiate the higher concentrations implies limitations imposed by clinical imaging techniques. Our results suggest that additional factors, such as paramagnetic material and motion as well as differences in protein concentration, may be involved in the MR signal intensities observed in pathologic CSF and cystic CNS collections

The T2 and pseudodensity (proportional to proton density) of intracranial hemorrhages and normal white matter were calculated. The mean T2 (+/- standard deviation) was 120 +/- 62 for hemorrhage and 61 +/- 11 for white matter. Pseudodensity values were normalized to a white matter value of 1, and the value for hemorrhage was 1.56 +/- 0.28. These values were used to determine which components of hemorrhage-white matter contrast are due to T1, T2, and density. The results indicate that on spin-echo (SE) images obtained with a long repetition time (TR)/short echo time (TE) (2,500/0-20 [TR msec/TE msec]), the contrast is mainly due to density differences, with a modest T2 contribution on 20-msec-TE images and nearly no T1 component. At 600/0-20, the contrast continues to be largely determined by density differences, again with a modest T2 component on 20-msec-TE images. If the T1 of hemorrhage is extremely short, the T1 component of contrast on 600/0-20 SE images will be somewhat greater than the density component. Because contrast on short TR/short TE images may be largely or entirely determined by pseudodensity or T2, it is inaccurate to refer to 600/20 images as 'T1-weighted'. The assumption that high signal intensity at this sequence implies a 'short T1' will lead to misleading conclusions

Magnetic resonance (MR) images of the lumbar spine from 150 patients were retrospectively reviewed. In 14 of these patients, at 18 disk levels, a vacuum phenomenon (VP) had been identified on plain radiographs and/or computed tomographic scans. The MR imaging appearance of these gas collections in 17 disks was an area without signal, best seen on spin-echo sequences with short repetition time and echo time in the sagittal view. MR imaging precisely located the VP in the anulus fibrosus, the nucleus pulposus, and Schmorl nodes. In all but one case, degeneration of the disk was complete and associated with adjacent changes in vertebral bone. Pitfalls of MR imaging detection of VP included chemical shift artifact, calcifications, and tears without gas in the disk

This is a report of the magnetic resonance study in a case of orbital metastasis from a carcinoid tumor. A rather unusual feature of this lesion was its relatively short T2 as compared with other metastatic orbital lesions

Thirteen orbital lesions in 12 patients were evaluated with both conventional spin-echo magnetic resonance (MR) imaging and phase-dependent proton spectroscopic imaging. This technique, which makes use of small differences in the resonant frequencies of water and fat protons, provides excellent high-resolution images with simultaneous chemical shift information. In this method, there is 180 degrees opposition of phase between fat protons and water protons at the time of the gradient echo, resulting in signal cancellation in voxels containing equal signals from fat and water. In this preliminary series, advantages of spectroscopic images in orbital lesions included better lesion delineation, with superior anatomic definition of orbital apex involvement; more specific characterization of high-intensity hemorrhage with a single pulse sequence; elimination of potential confusion from chemical shift misregistration artifact; further clarification of possible intravascular flow abnormalities; and improved apparent intralesional contrast

Fifty-nine cases in which surface coil MR imaging of the orbit was performed were reviewed. MR imaging was performed with spin-echo techniques at 1.5 T with both short repetition time/echo time (TR/TE) and long TR/TE sequences in all cases. In all patients short TR/TE images were obtained with small-diameter surface coils; long TR/TE images were usually obtained with a standard head coil. Surface coil MR appears to be an important adjunct in state-of-the-art orbital imaging. Orbital MR imaging may be most useful, providing information not available on computed tomography (CT), in identifying lesions in the orbital apex, superior orbital fissure, and optic canal; differentiating inflammatory pseudotumor from malignancy in clinically similar patients; characterizing lesions containing hemorrhage or other paramagnetic material; defining the posterior extent of optic pathway gliomas; and detecting abnormal flow in intraorbital vascular structures. CT seems to be superior to MR imaging in the evaluation of small perioptic meningiomas, especially those that are calcified