Rationale and objectives: To evaluate the ability of diffusion-weighted magnetic resonance imaging (DWI) in differentiating malignant thyroid nodules from benign lesions with a meta-analysis.
Materials and methods: Articles in English and Chinese language relating to the accuracy of DWI for this utility were retrieved. Pooled estimation and subgroup analysis data were obtained by statistical analysis.
Results: A total of seven studies (17 subsets) with 358 patients, who fulfilled all of the inclusion criteria, were considered for the analysis. No publication bias was found (bias = 7.03, P > .05). Methodological quality was relatively high. DWI sensitivity was 0.91 (95% confidence interval [CI], 0.87-0.94) and specificity was 0.93 (95% CI, 0.86-0.96). Overall, positive likelihood ratio was 12.24 (95% CI, 6.47-23.20) and negative likelihood ratio was 0.99 (95% CI, 0.06-0.15). Diagnostic odds ratio was 123.78 (95% CI, 56.85-269.48). The area under the curve of the summary receiver operating characteristic was 0.94 (95% CI, 0.92-0.96). In patients with high pretest probabilities, DWI enabled confirmation of malignant thyroid lesion; in patients with low pretest probabilities, DWI enabled exclusion of malignant thyroid lesion. Worst-case-scenario (pretest probability, 50%) posttest probabilities were 92% and 9% for positive and negative DWI results, respectively.
Conclusions: A limited number of small studies suggests that quantitative DWI is a reliable diagnostic method for differentiation between benign and malignant thyroid lesions.
Keywords: Diffusion-weighted imaging; meta-analysis; sensitivity; specificity; thyroid nodule.
Copyright © 2014 AUR. Published by Elsevier Inc. All rights reserved.