Radiomics Model for Predicting TP53 Status Using CT and Machine Learning Approach in Laryngeal Squamous Cell Carcinoma

• DOI: 10.3389/fonc.2022.823428
• Publication Date: 2022-04-28T05:03:18Z
• Authors: Ruxian Tian; Yumei Li; Chuanliang Jia; Yakui Mou; Haicheng Zhang; Xinxin Wu; Jingjing Li; Guohua Yu; Ning Mao; Xicheng Song
• Keywords: Lasso

Objective We aim to establish and validate computed tomography (CT)-based radiomics model for predicting TP53 status in patients with laryngeal squamous cell carcinoma (LSCC). Methods divided all into a training set 1 (n=66) testing (n=30) predict TP53. Radiomics features were selected by analysis of variance (ANOVA) the least absolute shrinkage selection operator (Lasso) regression analysis. Five models established using K-Nearest Neighbor, logistics regressive, linear-support vector machine (SVM), gaussian-SVM, polynomial-SVM 1. also 2 according different CT equipment evaluate stability models. Results After ANOVA subsequent Lasso analysis, 22 build The based on linear-SVM has best predictive performance five models, area under receiver operating characteristic curve 0.831(95% confidence interval [CI] 0.692–0.970) 0.797(95% CI 0.632–0.957) respectively. specificity, sensitivity, accuracy 0.971(95% 0.834–0.999), 0.714(95% 0.535–0.848), 0.843(95% 0.657–0.928) 0.750(95% 0.500–0.938), 0.786(95% 0.571–1.000), 0.667(95% 0.467–0.720) 1, In addition, achieved stable prediction results even equipment. Decision showed that could benefit LSCC patients. Conclusion developed validated relatively optimal trying learning methods LSCC. It shown great potential preoperatively guiding clinical treatment.