Objective CT criteria to determine the presence of abnormal basal enhancement in children with suspected tuberculous meningitis

Date
2006
Authors
Przybojewski S.
Andronikou S.
Wilmshurst J.
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Background: There are no widely accepted objective criteria to determine the presence of basal enhancement on CT in children with suspected tuberculous meningitis (TBM). Objective: To test nine recently described objective CT criteria for the presence of abnormal basal enhancement in children with suspected TBM against the definite diagnosis as determined by cerebrospinal fluid (CSF) culture. Materials and methods: CT scans of patients with a clinical suspicion of TBM who had undergone lumbar puncture for CSF culture spanning a period of 4 years were reviewed for the presence of nine recently described criteria for the presence of abnormal basal enhancement. The radiologists were blinded to the final diagnosis based on CSF culture against which the criteria were tested. The criteria have been named: the 'Y-sign', 'linear enhancement', 'double lines', Iinfundibular recess of the third', 'ill-defined edge', 'nodular enhancement', 'join the dots', 'contrast filling the cisterns', and 'asymmetry'. Results: A total of 65 patients were included in the study, 34 with culture-proven TBM and 31 with other diagnoses. Four individual criteria had a specificity of 100%, but the sensitivities of these criteria ranged from 15% to 53% only. Three other criteria had specificities of 97% and sensitivities ranging from 62% to 82%. The presence of more than one criterion in the same patient showed a specificity of 97% and sensitivity of 91%. Conclusions: Very high specificity was demonstrated for all nine criteria, including 100% specificity for four individual criteria. Sensitivity was at best 82%, but improved to 91% when more than one criterion was present. These criteria need to be tested for inter- and intraobserver variability to prove their clinical usefulness. © Springer-Verlag 2006.
Description
Keywords
contrast medium, cerebrospinal fluid culture, child, computer assisted tomography, contrast enhancement, controlled study, diagnostic accuracy, diagnostic value, female, gold standard, human, image analysis, image display, infant, intermethod comparison, lumbar puncture, major clinical study, male, observer variation, priority journal, radiodiagnosis, review, roc curve, sensitivity and specificity, tuberculous meningitis, validation process, Child, Child, Preschool, Contrast Media, Diagnosis, Differential, Female, Humans, Infant, Male, Sensitivity and Specificity, Spinal Puncture, Tomography, X-Ray Computed, Tuberculosis, Meningeal
Citation
Pediatric Radiology
36
7