Can Diffusion Weighted Imaging Aid in Differentiating Benign from Malignant Sinonasal Masses?: A Useful Adjunct
More details
Hide details
Publication date: 2018-02-02
Pol J Radiol, 2017; 82: 345-355
Background: To evaluate the role of diffusion weighted imaging (DWI) and apparent diffusion coefficient (ADC) values at 3 Tesla in characterizing sinonasal masses.
Material/Methods: After ethical clearance, 79 treatment naive patients with head and neck masses underwent magnetic resonance imaging (MRI), including DWI at 3 Tesla using the following b values – 0, 500 and 1000 s/mm2. Thirty-one patients were found to have sinonasal tumours and were subsequently analysed. Image analysis consisted of a morphological evaluation of conventional MR images, qualitative evaluation of DW trace images and quantitative assessment of mean ADC values. Receiver operating characteristic (ROC) curve was drawn to determine a cut-off ADC value for the differentiation between benign and malignant masses.
Results: Sinonasal masses showed an overlapping growth pattern on conventional imaging, irrespective of their biological nature. The mean ADC value of benign lesions was 1.948±0.459×10–3 mm2/s, while that of malignant lesions was 1.046±0.711×10–3 mm2/s, and the difference was statistically significant (p=0.004). When a cut-off ADC value of 1.791×10–3 mm2/s was used, sensitivity of 80% and specificity of 83.3% were obtained for characterization of malignant lesions, which was statistically significant. Juvenile nasopharyngeal angiofibroma (JNA) showed distinctly high ADC values, while meningioma was the only benign lesion with restricted diffusion. Atypical entities with unexpected diffusion characteristics included: adenocarcinoma, adenoid cystic carcinoma, meningioma, chondrosarcoma and fibromyxoid sarcoma.
Conclusions: DWI in conjunction with conventional imaging can potentially enhance the diagnostic accuracy in characterizing sinonasal masses as benign or malignant. Some specific entities such as JNA and meningioma showed distinctive diffusion characteristics.
Journals System - logo
Scroll to top