Papillary Thyroid Cancer: Dual-Energy Spectral CT Quantitative Parameters for Preoperative Diagnosis of Metastasis to the Cervical Lymph Nodes

PURPOSE: To evaluate the use of dual-energy spectral computed tomographic (CT) quantitative parameters compared with the use of conventional CT imaging features for preoperative diagnosis of metastasis to the cervical lymph nodes in patients with papillary thyroid cancer. MATERIALS AND METHODS: This study was approved by the ethics committee and all patients provided written informed consent. Analyses of quantitative gemstone spectral imaging data and qualitative conventional CT imaging features were independently performed by different groups of radiologists. Excised lymph nodes were located and labeled during surgery according to location on preoperative CT images and were evaluated histopathologically. Single and combined parameters were fitted to simple and multiple logistic regression models, respectively, by means of the generalized estimating equations method. Sensitivity and specificity analyses were performed by using receiver operating characteristic curves and were compared with data from the qualitative analysis. RESULTS: The slope of the spectral Hounsfield unit curve ( λHU slope of the spectral Hounsfield unit curve ), normalized iodine concentration, and normalized effective atomic number measured during both arterial and venous phases were significantly higher in metastatic than in benign lymph nodes. The best single parameter for detection of metastatic lymph nodes was venous phase λHU, with sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of 62.0%, 91.1%, 80.6%, 79.7%, and 81.0%, respectively. The best combination of parameters was venous phase λHU slope of the spectral Hounsfield unit curve and arterial phase normalized iodine concentration, with values of 73.0%, 88.4%, 82.9%, 78.0%, and 85.3%, respectively. Compared with qualitative analysis, the venous phase λHU slope of the spectral Hounsfield unit curve showed higher specificity (91.1% vs 83.0%, P < .001) and similar sensitivity (62.0% vs 61.9%, P > .99), and the combined venous phase λHU slope of the spectral Hounsfield unit curve and arterial phase normalized iodine concentration showed higher sensitivity (73.0% vs 61.9%, P < .001) and specificity (88.4% vs 83.0%, P < .001). CONCLUSION: Quantitative assessment with gemstone spectral imaging quantitative parameters showed higher accuracy than qualitative assessment of conventional CT imaging features for preoperative diagnosis of metastatic cervical lymph nodes in patients with papillary thyroid cancer.