Ultrasonographic Multimodality Diagnostic Model of Thyroid Nodules.

Abstract

The aim of this study was to identify independent risk factors for thyroid cancer, establish an ultrasonographic multimodality diagnostic model for thyroid nodules, and explore the diagnostic value of the model. From November 2011 to February 2015, 307 patients with a total of 367 thyroid nodules underwent conventional ultrasound, contrast-enhanced ultrasound (CEUS), and ultrasound elastography examinations before surgery. A binary logistic regression analysis was performed to identify independent risk factors for thyroid cancer and to establish a multimodality diagnostic model for thyroid nodules. The diagnostic performance of conventional ultrasound, CEUS, ultrasound elastography, and the multimodality diagnostic model was assessed and compared. The following seven independent risk factors were included in the logistic regression models: age, irregular shape, hypoechoic pattern, marked hypoechoic pattern, irregular blood flow distribution, heterogeneous enhancement, and an elastic score of 3/4. The multimodality diagnostic model had a diagnostic accuracy of 86.9%, with a sensitivity of 93.5% and a specificity of 77.3%. The multimodality diagnostic model improved the diagnostic accuracy compared with that of conventional ultrasound, CEUS, and ultrasound elastography. Independent risk factors for thyroid cancer included age, irregular shape, hypoechoic pattern, marked hypoechoic pattern, irregular blood flow distribution, heterogeneous enhancement, and an elastic score of 3/4. The multimodality diagnostic model was demonstrated to be effective in the diagnosis of thyroid nodules.