Differentiation of Papillary Thyroid Microcarcinoma and Benign Thyroid Nodules Using CT Diameter Ratio
10.3969/j.issn.1005-5185.2015.11.006
- VernacularTitle:CT径线比鉴别甲状腺微小癌与良性结节的价值
- Author:
Wenhui DAI
;
Liping CHAI
;
Lixin SUN
;
Jianchang CHEN
;
Haiyan FU
;
Li WANG
;
Xuewen YU
;
Shanfeng LIU
- Publication Type:Journal Article
- Keywords:
Thyroid neoplasms;
Thyroid nodule;
Tomography,spiral computed;
Image processing,computer-assisted;
ROC curve;
Diagnosis,differential
- From:
Chinese Journal of Medical Imaging
2015;(11):819-823
- CountryChina
- Language:Chinese
-
Abstract:
Purpose Papillary thyroid microcarcinoma (PTMC) is difficult to diagnose its nature before surgery, thus results in misdiagnosis. This paper aims to determine the best diagnostic cutoff value using anteroposterior and transverse diameter ratio (A/L) and longitudinal and transverse diameter ration (L/T) in PTMC. Materials and Methods The CT data of 154 pathology proven benign and malignant thyroid nodules ≤ 1.0 cm in diameter in 78 cases were reviewed, including 75 PTMC in 47 patients and 79 benign nodule in 31 patients. The anteroposterior and transverse diameter ratio (A/T) on axial view, A/L on sagittal view, and L/T on coronal view were measured and calculated. A non-parametric method was used to draw the receiver operating curve of A/T, A/L and L/T. The mean and standard deviation of CT diameters in benign and malignant nodules were calculated. The area under the curve, sensitivity, specificity and diagnostic accuracy, positive predictive value and negative predictive value were determined. CT manifestations of small benign and malignant thyroid nodules were also analyzed. Results The A/T and A/L ratio were significantly larger for PTMC than benign nodules (P<0.01), while L/T diameters were significantly smaller than the benign nodules (P<0.01). The area under ROC was 0.8841, 0.7676 and 0.4052 for A/T, A/L and L/T respectively. The best diagnostic cutoff value of A/T and A/L were 1.05 and 1.0. With A/T ≥ 1.05, the sensitivity, specificity, accuracy, positive predictive value and negative predictive value were 88.00%, 84.81%, 86.36%, 84.62% and 88.12%, respectively. With A/L ≥ 1.0, the sensitivity, specificity, accuracy, positive predictive value and negative predictive values were 66.67%, 82.28%, 74.68%, 78.13% and 72.22%, respectively. CT characteristics of PTMC included superficial location, oval low density, blurry boundary, microcalcification and progressive enhancement with accuracy of 71.43%, 50.00%, 79.22%, 68.83% and 90.91%, respectively. Conclusion The cutoff values of A/T ≥ 1.05 and A/L ≥ 1.0 can be used to diagnose PTMC with high accuracy.
