二维超声图像特征联合剪切波弹性成像预测甲状腺乳头状癌BRAF V600E基因型OA

Objective To investigate the clinical utility of combining two-dimensional ultrasound image features with shear wave elastography (SWE) parameters in predicting mutations of the B-Raf proto-oncogene,serine/threonine kinase,Val600Glu(BRAFV600E) in patients with papillary thyroid carcinoma (PTC). Methods A total of 120 patients with PTC who underwent thyroid ultrasound examination at Xingtai People's Hospital from December 2024 to May 2025 and obtained postoperative BRAF V600E genetic testing results were selected as study subjects. The two-dimensional ultrasound features and SWE characteristics were analyzed,feature screening was performed using L1-regularized Logistic regression,and two-dimensional ultrasound models,elastic imaging models,and combined models were constructed. Model performance was evaluated employing machine learning algorithms. Results Under the multi-layer perceptron algorithm,the area under the curve of the receiver operating characteristics in the joint model reached 0.902,significantly outperforming that of the single-modal model. The BRAF mutation group exhibited a higher proportion of solid nodules,a long-to-wide ratio>1,irregular margins,microcalcifications,and elevated elastographic quality index values (P<0.05). Conclusion The combination of conventional ultrasound features and SWE parameters can significantly improve the preoperative predictive performance for BRAFV600E mutation in PTC,providing a reference for personalized diagnosis and treatment.