Objective: To explore the utility of lung magnetic resonance imaging (MRI) in children with thoracic tuberculosis (TB).
Methods: This prospective study of forty children (age range of 5 to 15 y) with thoracic TB was approved by the institutional ethics committee. Chest radiograph, lung MRI and computed tomography (CT) scan were performed in all children. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) of MRI were calculated and kappa test of agreement was used to determine the strength of agreement between the MRI and CT findings.
Results: MRI performed equivalent to CT scan in detection of mediastinal and hilar lymphadenopathy, pleural effusion and lung cavitation (considered typical for TB) with sensitivity and specificity of 100%. MRI also yielded a sensitivity of 88.2% and specificity of 95.7% for nodules, with sensitivity of 100% and specificity of 92.9% for consolidation. Kappa agreement between CT and MRI in detection of each finding was almost perfect (k: 0.8-1).
Conclusions: Although CT scan is still superior to MRI in detection of smaller nodules, MRI demonstrates a very high degree of correlation and agreement with CT scan for detecting lung and mediastinal pathology related to TB, in children with thoracic TB and can be particularly useful in select population and follow-up of these children to avoid repeated radiation exposures.
Keywords: CT scan; Children; Imaging; MRI; Tuberculosis; X-ray.