Multiparametric MR imaging for differentiating between benign and malignant thyroid nodules: Initial experience in 23 patients

Abstract

To evaluate multiparametric MR imaging with combined use of apparent diffusion coefficients (ADCs) and time-signal intensity curves (TICs) for discriminating malignant thyroid nodules from benign ones. ADCs and TICs of 7 benign and 16 malignant nodules were retrospectively analyzed on lesion-by-lesion (overall ADCs and TICs) and pixel-by-pixel (ADC and TIC mapping) bases. ADCs were determined using b-values of 500 and 1000 s/mm(2) . The TICs were classified into 5 patterns on the basis of increment ratio, peak time, and washout ratio. Stepwise approach based on ADC and TIC criteria was used to discriminate between benign and malignant nodules. Overall ADC discriminated undifferentiated carcinomas from papillary carcinomas at 91% accuracy (≤1.3 × 10(-3) mm(2) /s) and differentiated lymphomas from the other malignant nodules at 100% accuracy (≤0.65 × 10(-3) mm(2) /s). Most malignant thyroid nodules had large (≥45%) areas of rapid-uptake TIC profile with small (≤15%) areas of flat TIC profile, or had small (<45%) areas of rapid-uptake TIC profile with large (≥25%) areas of extremely low or low ADCs (≤1.2 × 10(-3) mm(2) /s). Stepwise approach discriminated malignant nodules from benign ones at 91% accuracy. The multiparametric MR imaging helps discriminating malignant thyroid nodules from benign ones.