Prediction of contralateral central lymph node metastasis in unilateral papillary thyroid carcinoma based on radiomics.

This study aims to examine the determinants of contralateral central lymph node metastasis in cases of unilateral papillary thyroid carcinoma, utilizing clinical pathological parameters and ultrasound radiomics. The goal is to develop an optimal predictive model that can inform clinical decisions regarding contralateral central lymph node dissection in patients with unilateral papillary thyroid cancer. A cohort of 329 patients diagnosed with unilateral papillary thyroid carcinoma, all of whom underwent bilateral thyroidectomy and bilateral central lymph node resection, were analyzed. Clinical data were systematically collected, and a logistic regression analysis model was constructed. The patient cohort was randomly divided into a training set and a validation set in an 8:2 ratio. Radiomic features were extracted from ultrasound images using the open-source software pyradiomic, leading to the establishment of three models: a clinical model, a radiomics model, and a combined clinical and radiomics model. The area under the curve (AUC) for the logistic regression model was found to be 0.843. In the training set, the AUCs for the clinical model, the radiomics feature model, and the combined clinical and radiomics feature model developed using random forest were 0.7645, 0.9633, and 0.9726, respectively. In the validation set, the AUCs for the clinical model, the radiomics feature model, and the combined clinical and radiomics feature model developed using random forest were 0.7358, 0.9558, and 0.9694, respectively. The combined model incorporated four clinical features and seven radiomic features, specifically: age, tumor size, presence of microcalcification, presence of ipsilateral central lymph node metastasis, elongation, perimeter, sphericity, maximum probability, large dependence emphasis, first-order range, and cluster shade. The combined clinical and radiomics model developed in this research demonstrates strong predictive diagnostic efficacy. The establishment of this innovative model is anticipated to provide a robust theoretical framework for clinicians considering preventive bilateral central lymph node dissection in patients with unilateral papillary thyroid carcinoma.