White matter lesions by magnetic resonance: an assessment between DWI and T2 FLAIR techniques

Objectives – To determine the sensitivity and specificity of diffusion weighted (DWI) and Fluid-Attenuated Inversion Recovery (T2 FLAIR) in the evaluation of white matter lesions and to verify what is his extended complement in order to create a set of best practices in the MRI routine of the brain. Methodology – Was carried out a quantitative study with a retrospective analysis. From database, were selected 30 patients, 20 with and 10 without disease. We obtained a sample of 60 images, including 30-weighted images of T2 FLAIR and 30 of DWI. Using the program Viewdex®, the set of images were observed by three observers. The images were assessed according seven criteria: visibility, detection, consistency, location, size and margins of the lesion and diagnostic capability. We used the calculation of sensitivity and specificity by ROC curves, as well as statistical analysis, namely, T-test, Kappa index and the Pearson’s correlation coefficient between the variables under study. Results – The sensitivity and specificity obtained for the weighted T2 FLAIR were higher (0.915 and 0.038 respectively) to the weighting DWI (0.08 and 0.100 respectively). There were no significant population variances. We obtained a high linear correlation between variables with an r value located between 0.8 and 0.99. There was a considerable variability among observers. Conclusions – Although the values ofsensitivity and specificity for DWI have been lower than expected, we suggest that this sequence should be included in the routine protocol of the brain to help the differential diagnosis with other pathologies.