Search

The term rhombencephalitisrefers to inflammatory diseases affecting the hindbrain (brainstem and cerebellum). Rhombencephalitis has a wide variety of etiologies, some of them potentially severe and life threatening without proper early treatment. In this retrospective observational study, we reviewed the records of consecutively hospitalized patients at University Hospital of Bellvitge (Barcelona, Spain) from January 1990 to December 2008. Rhombencephalitis was defined as a brainstem and/or cerebellar condition demonstrated clinically or by neuroimaging, with pleocytosis (>4 cells/mm3) in cerebrospinal fluid. Ninety-seven patients (48 female; mean age, 37 yr; range, 14-79 yr) fulfilled these criteria. We reviewed their clinical, cerebrospinal fluid, and radiologic characteristics. The mean follow-up was 5 years (range, 0-20 yr). The etiologies of rhombencephalitis were as follows unknown cause (n = 31), multiple sclerosis (n = 28), Behçet disease (n = 10), Listeria monocytogenesinfection (n = 9), paraneoplastic syndrome (n = 6) (3 cases associated with anti-Yo antibodies and 3 with anti-Tr antibodies), Epstein-Barr virus (n = 4), tuberculosis (n = 2), pneumococcal infection (n = 2), systemic lupus erythematosus (n = 1), lymphoma (n = 1), Brucella speciesinfection (n = 1), JC virus (n = 1), and relapsing polychondritis (n = 1). Certain clinical, cerebrospinal fluid, and radiologic characteristics that are commonly seen in some of these etiologies can guide us in the first approach to the etiologic diagnosis of rhombencephalitis.

Automated pattern recognition techniques are needed to help radiologists categorize MRS data of brain tumors according to histological type and grade. A major question is whether a computer program “trained” on spectra from one hospital will be able to classify those from another, particularly if the acquisition protocol is different. A subset of 144 histopathologically validated brain tumor spectra in the INTERPRET database, obtained from three of the collaborating centers, was grouped into meningiomas, low-grade astrocytomas, and “aggressive tumors” (glioblastomas and metastases). Spectra from two centers formed the training set (94 spectra) while the third acted as the test set (50 spectra). Linear discriminant analysis successfully classified 48/50 in the test set; the remaining two were atypical cases. When the training and test sets were combined, 133 of the 144 spectra were correctly classified using the leave-one-out procedure. These spectra had been obtained using different sequences (STEAM and PRESS), different echo times (20, 30, 31, and 32 ms), different repetition times (1600 and 2000 ms), and different manufacturers' instruments (GE and Philips). Pattern recognition algorithms are less sensitive to acquisition parameters than had been expected. Magn Reson Med 49:29–36, 2003. © 2003 Wiley-Liss, Inc.

Automated pattern recognition techniques are needed to help radiologists categorize MRS data of brain tumors according to histological type and grade. A major question is whether a computer program “trained” on spectra from one hospital will be able to classify those from another, particularly if the acquisition protocol is different. A subset of 144 histopathologically validated brain tumor spectra in the INTERPRET database, obtained from three of the collaborating centers, was grouped into meningiomas, low‐grade astrocytomas, and “aggressive tumors” (glioblastomas and metastases). Spectra from two centers formed the training set (94 spectra) while the third acted as the test set (50 spectra). Linear discriminant analysis successfully classified 48/50 in the test set; the remaining two were atypical cases. When the training and test sets were combined, 133 of the 144 spectra were correctly classified using the leave‐one‐out procedure. These spectra had been obtained using different sequences (STEAM and PRESS), different echo times (20, 30, 31, and 32 ms), different repetition times (1600 and 2000 ms), and different manufacturers' instruments (GE and Philips). Pattern recognition algorithms are less sensitive to acquisition parameters than had been expected. Magn Reson Med 49:29–36, 2003. © 2003 Wiley‐Liss, Inc.