Head and Neck Lesions: Characterization with Diffusion-weighted Echo-planar MR Imaging

Abstract

PURPOSE: To evaluate whether apparent diffusion coefficients (ADCs) calculated from diffusion-weighted echo-planar magnetic resonance (MR) images can be used to characterize head and neck lesions. MATERIALS AND METHODS: Diffusion-weighted echo-planar MR imaging was performed with a 1.5-T MR unit in 97 head and neck lesions in 97 patients. Images were obtained with a diffusion-weighted factor, factor b, of 0, 500, and 1,000 sec/mm(2), and an ADC map was constructed. The ADCs of lesions, cerebrospinal fluid, and spinal cord were calculated. RESULTS: Acceptable images for ADC measurement were obtained in 81 (84%) patients. The mean ADC of malignant lymphomas, (0.66 +/- 0.17[SD]) x 10(-3) mm(2)/sec (n = 13), was significantly smaller (P <.001) than that of carcinomas. The mean ADC of carcinomas, (1.13 +/- 0.43) x 10(-3) mm(2)/sec (n = 36), was significantly smaller (P =.002) than that of benign solid tumors. The mean ADC of benign solid tumors, (1.56 +/- 0.51) x 10(-3) mm(2)/sec (n = 22), was significantly smaller (P =.035) than that of benign cystic lesions, (2.05 +/- 0.62) x 10(-3) mm(2)/sec (n = 10). No significant differences were seen in the mean ADC of cerebrospinal fluid and of spinal cord among four groups of lesions. When an ADC smaller than 1.22 x 10(-3) mm(2)/sec was used for predicting malignancy, the highest accuracy of 86%, with 84% sensitivity and 91% specificity, was obtained. CONCLUSION: Measurement of ADCs may be used to characterize head and neck lesions.