Histogram-based analysis of diffusion-weighted imaging for predicting aggressiveness in papillary thyroid carcinoma

Abstract

Abstract Background To assess the potential of apparent diffusion coefficient (ADC) map in predicting aggressiveness of papillary thyroid carcinoma (PTC) based on whole-tumor histogram-based analysis. Methods A total of 88 patients with PTC confirmed by pathology, who underwent neck magnetic resonance imaging, were enrolled in this retrospective study. Whole-lesion histogram features were extracted from ADC maps and compared between the aggressive and non-aggressive groups. Multivariable logistic regression analysis was performed for identifying independent predictive factors. Receiver operating characteristic curve analysis was used to evaluate the performances of significant factors, and an optimal predictive model for aggressiveness of PTC was developed. Results The aggressive and non-aggressive groups comprised 67 (mean age, 44.03 ± 13.99 years) and 21 (mean age, 43.86 ± 12.16 years) patients, respectively. Five histogram features were included into the final predictive model. ADC_firstorder_TotalEnergy had the best performance (area under the curve [AUC] = 0.77). The final combined model showed an optimal performance, with AUC and accuracy of 0.88 and 0.75, respectively. Conclusions Whole-lesion histogram analysis based on ADC maps could be utilized for evaluating aggressiveness in PTC.