This retrospective study aimed to evaluate whether ultrasound texture analysis is useful to predict lymph node metastasis in patients with papillary thyroid microcarcinoma (PTMC). This study was approved by the Institutional Review Board, and the need to obtain informed consent was waived. Between May and July 2013, 361 patients (mean age, 43.8 ± 11.3 years; range, 16-72 years) who underwent staging ultrasound (US) and subsequent thyroidectomy for conventional PTMC ≤ 10 mm between May and July 2013 were included. Each PTMC was manually segmented and its histogram parameters (Mean, Standard deviation, Skewness, Kurtosis, and Entropy) were extracted with Matlab software. The mean values of histogram parameters and clinical and US features were compared according to lymph node metastasis using the independent t-test and Chi-square test. Multivariate logistic regression analysis was performed to identify the independent factors associated with lymph node metastasis. Tumors with lymph node metastasis (n = 117) had significantly higher entropy compared to those without lymph node metastasis (n = 244) (mean±standard deviation, 6.268±0.407 vs. 6.171±.0.405; P =.035). No additional histogram parameters showed differences in mean values according to lymph node metastasis. Entropy was not independently associated with lymph node metastasis on multivariate logistic regression analysis (Odds ratio, 0.977 [95% confidence interval (CI), 0.482-1.980]; P =.949). Younger age (Odds ratio, 0.962 [95% CI, 0.940-0.984]; P =.001) and lymph node metastasis on US (Odds ratio, 7.325 [95% CI, 3.573-15.020]; P <.001) were independently associated with lymph node metastasis. Texture analysis was not useful in predicting lymph node metastasis in patients with PTMC.
Bibliographical notePublisher Copyright:
© 2017 Kim et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which\ permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
All Science Journal Classification (ASJC) codes