K R Davis

In 41 cases of verified ruptured saccular aneurysm, we prospectively predicted the presence or absence of delayed symptomatic cerebral vasospasm. CT criteria quantifying the extent and location of subarachnoid blood (developed in our previous retrospective study) were used in this prospective series of patients. Twenty-two patients had recognizable subarachnoid clots larger than 3 X 5 mm or layers of blood more than 1 mm thick (measured on reproduced images). In 20 of the 22 patients with severe significant clot or thick layer, severe vasospasm was correctly predicted and localized (2 false positives). In 19 patients with no blood, or diffuse blood, or blood outside the subarachnoid space, the absence of severe vasospasm was correctly predicted in 14 (5 false negatives). All of the false-positive and false-negative cases could be explained by inadequate CT technique. The data indicate that the extent and location of blood in the subarachnoid space determine the severity and location of vasospasm and that patients in jeopardy of developing symptomatic cerebral vasospasm can now be identified. Early preventive measures may now be assessed more accurately.

The dynamic signal intensity changes at magnetic resonance (MR) imaging in active and chronic wallerian degeneration in the corticospinal tract were evaluated. Forty-three patients with wallerian degeneration seen on MR images after cerebral infarction were studied. When possible, patients with acute stroke were examined with MR imaging prospectively at the onset of symptoms and then at weekly intervals for several months. Focal infarction without distal axonal degeneration is demonstrated for the 1st month following onset of clinical symptoms. At 4 weeks, a well-defined band of hypointense signal appears on T2-weighted images in the topographic distribution of the corticospinal tract. After 10-14 weeks, the signal becomes permanently hyperintense. Over several years, accompanying ipsilateral brain stem shrinkage occurs. The dark signal intensity observed on T2-weighted images between 4 and 14 weeks is believed to result primarily from transitory increased lipid-protein ratio.

We report our experience with the neurologic sequelae (at a mean follow-up of 24 months) among the 15 surviving infants who have had neonatal intraventricular hemorrhage (IVH) documented by computerized tomographic (CT) brain scan. Neurologically six infants (40%) are normal, six infants (40%) mildly impaired, and three infants (20%) moderate to severely impaired. The neurologic outcome correlated to the degree of hemorrhage seen in the CT scans when IVH was classified into four grades. None of the other neonatal factors examined showed significant correlation with the outcome.

The authors investigated whether identification of corpus callosal (CC) involvement might increase the specificity of magnetic resonance (MR) imaging in differentiating multiple sclerosis (MS) from other periventricular white matter diseases (PWDs). They prospectively evaluated 42 patients with MS and 127 control patients with other PWDs. Ninety-three percent of the MS patients demonstrated confluent and/or focal lesions involving the callosal-septal interface (CSI). These lesions characteristically involved the inferior aspect of the callosum and radiated from the ventricular surface into the overlying callosum. CSI lesions were optimally demonstrated on sagittal long repetition time (TR)/short echo time (TE) images and frequently (45% of cases) went undetected on axial images. Only 2.4% of the control patients had lesions of the CC. The authors conclude that midsagittal long TR/short TE images are highly sensitive and specific for MS and that callosal involvement in MS is more common than previously reported.

Using a modification of the partial saturation (PS) pulse sequence, we developed an MR method that permits the acquisition of highly T1- and T2-weighted images of the head and body in as little as 10 sec. The PS images, which were acquired at 0.6 T in a series of six patients with acute and subacute hemorrhage, showed a striking reduction in the signal intensity of hemorrhagic lesions. This effect, which is related to bulk magnetic susceptibility variations, was either minimal or absent on conventional T1- and T2-weighted spin-echo (SE) images. Our results suggest that high-field systems are not needed in order to image acute and subacute hemorrhage.