Radiomic Nomogram Improves Preoperative T Category Accuracy in Locally Advanced Laryngeal Carcinoma

作者：Fei Wang, Bin Zhang, Xiangjun Wu, Lizhi Liu, Fang Jin, Qiuying Chen, Minmin Li, Zhuozhi Chen, Yueyue Li, Di Dong, Jie Tian, Shuixing Zhang · 发表于：Frontiers in Oncology · 年份：2019 · DOI：10.3389/fonc.2019.01064 · 被引用次数：39 · 研究领域：Radiomics and Machine Learning in Medical Imaging、Head and Neck Cancer Studies、Lung Cancer Diagnosis and Treatment

Surgical decision-making on advanced laryngeal carcinoma is heavily depended on the identification of preoperative T category (T3 vs. T4), which is challenging for surgeons. A T category prediction radiomics (TCPR) model would be helpful for subsequent surgery. A total of 211 patients with locally advanced laryngeal cancer who had undergone total laryngectomy were randomly classified into the training cohort (n = 150) and the validation cohort (n = 61). We extracted 1390 radiomic features from the contrast-enhanced computed tomography images. Interclass correlation coefficient and the least absolute shrinkage and selection operator (LASSO) analyses were performed to select features associated with pathology-confirmed T category. Eight radiomic features were found associated with preoperative T category. The radiomic signature was constructed by Support Vector Machine algorithm with the radiomic features. We developed a nomogram incorporating radiomic signature and T category reported by experienced radiologists. The performance of the model was evaluated by the area under the curve (AUC). The T category reported by radiologists achieved an AUC of 0.775 (95% CI: 0.667-0.883); while the radiomic signature yielded a significantly higher AUC of 0.862 (95% CI: 0.772-0.952). The predictive performance of the nomogram incorporating radiomic signature and T category reported by radiologists further improved, with an AUC of 0.892 (95% CI: 0.811-0.974). Consequently, for locally advanc...