Nomogram Based On Contrast-enhanced Ultrasound Can Predict Malignancy In Thyroid Nodules

Background Thyroid nodules are a common disease.The correct diagnosis of thyroid nodules at an early stage has important clinical significance.Objective This study aimed to determine the feasibility of using a radiomics nomogram to predict thyroid carcinoma from contrast-enhanced ultrasound imagines.Methods Data set of US images in patients with thyroid nodules from First Affiliated Hospital,School of Medicine,Shihezi University,Tianyou Hospital Affiliated to Wuhan University of Science and Technology,Yichang Central People’s Hospital The First College of Clinical Medical,Affiliated Renhe Hospital of Three Gorges University and The First People’s Hospital of Yueyang(372 images from 2017 to 2019)were collected.They comprised the training and validation sets.Radiomics features were extracted from contrast enhanced ultrasound(CEUS)images and B-mode ultrasound(BMUS)images.Using the minimum redundancy maximum relevance algorithm and the least absolute shrinkage and selection operator regression to select thyroid cancer related features,and construct CEUS and BMUS radiomics score(Rad-score).Binary logistic regression was performed using two radiomics scores and clinically independent risk factors,and a nomogram was finally formulated.The nomogram’ performance was analyzed in terms of accuracy,sensitivity,specificity,receiver operating characteristic curves,and areas under the receiver operating characteristic curve(AUCs).Results BMUS and CEUS radiomics scores in patients with thyroid cancer were significantly higher than those in patients with benign thyroid nodules.Logistic regression analysis suggested that CEUS Rad score,TI-RADS level were independent risk factors related to thyroid cancer.The radiomics nomogram achieved an AUC of 0.987(95%confidence interval [CI]: 0.977,0.997)in the prediction of malignant thyroid nodules in the training set and the AUC in validation set was 0.995(95% confidence interval [CI]: 0.987,1).The C index of the training and validation set [0.987(0.978,0.996)vs0.996(0.990,1.002)] were both > 0.9,indicating that our CEUS radiomics nomogram has favorable discriminatory efficacy in thyroid cancer.Conclusion Using CEUS images from patients with thyroid cancer,radiomics nomogram can effectively predict malignant thyroid nodules.