Zhongyi Wang1,2, Fan Lin1,2, Heng Ma2, Yinghong Shi2, Jianjun Dong2, Ping Yang3, Kun Zhang4, Na Guo5, Ran Zhang5, Jingjing Cui5, Shaofeng Duan6, Ning Mao2, Haizhu Xie2. 1. School of Medical Imaging, Binzhou Medical University, Yantai, China. 2. Department of Radiology, Yantai Yuhuangding Hospital, Affiliated Hospital of Qingdao University, Yantai, China. 3. Department of Pathology, Yantai Yuhuangding Hospital, Affiliated Hospital of Qingdao University, Yantai, China. 4. Department of Breast Surgery, Yantai Yuhuangding Hospital, Affiliated Hospital of Qingdao University, Yantai, China. 5. Collaboration Department, Huiying Medical Technology Co., Ltd, Beijing, China. 6. Precision Health Institution, GE Healthcare, Shanghai, China.
Abstract
PURPOSE: We developed and validated a contrast-enhanced spectral mammography (CESM)-based radiomics nomogram to predict neoadjuvant chemotherapy (NAC)-insensitive breast cancers prior to treatment. METHODS: We enrolled 117 patients with breast cancer who underwent CESM examination and NAC treatment from July 2017 to April 2019. The patients were grouped randomly into a training set (n = 97) and a validation set (n = 20) in a ratio of 8:2. 792 radiomics features were extracted from CESM images including low-energy and recombined images for each patient. Optimal radiomics features were selected by using analysis of variance (ANOVA) and least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation, to develop a radiomics score in the training set. A radiomics nomogram incorporating the radiomics score and independent clinical risk factors was then developed using multivariate logistic regression analysis. With regard to discrimination and clinical usefulness, radiomics nomogram was evaluated using the area under the receiver operator characteristic (ROC) curve (AUC) and decision curve analysis (DCA). RESULTS: The radiomics nomogram that incorporates 11 radiomics features and 3 independent clinical risk factors, including Ki-67 index, background parenchymal enhancement (BPE) and human epidermal growth factor receptor-2 (HER-2) status, showed an encouraging discrimination power with AUCs of 0.877 [95% confidence interval (CI) 0.816 to 0.924] and 0.81 (95% CI 0.575 to 0.948) in the training and validation sets, respectively. DCA revealed the increased clinical usefulness of this nomogram. CONCLUSION: The proposed radiomics nomogram that integrates CESM-derived radiomics features and clinical parameters showed potential feasibility for predicting NAC-insensitive breast cancers.
PURPOSE: We developed and validated a contrast-enhanced spectral mammography (CESM)-based radiomics nomogram to predict neoadjuvant chemotherapy (NAC)-insensitive breast cancers prior to treatment. METHODS: We enrolled 117 patients with breast cancer who underwent CESM examination and NAC treatment from July 2017 to April 2019. The patients were grouped randomly into a training set (n = 97) and a validation set (n = 20) in a ratio of 8:2. 792 radiomics features were extracted from CESM images including low-energy and recombined images for each patient. Optimal radiomics features were selected by using analysis of variance (ANOVA) and least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation, to develop a radiomics score in the training set. A radiomics nomogram incorporating the radiomics score and independent clinical risk factors was then developed using multivariate logistic regression analysis. With regard to discrimination and clinical usefulness, radiomics nomogram was evaluated using the area under the receiver operator characteristic (ROC) curve (AUC) and decision curve analysis (DCA). RESULTS: The radiomics nomogram that incorporates 11 radiomics features and 3 independent clinical risk factors, including Ki-67 index, background parenchymal enhancement (BPE) and human epidermal growth factor receptor-2 (HER-2) status, showed an encouraging discrimination power with AUCs of 0.877 [95% confidence interval (CI) 0.816 to 0.924] and 0.81 (95% CI 0.575 to 0.948) in the training and validation sets, respectively. DCA revealed the increased clinical usefulness of this nomogram. CONCLUSION: The proposed radiomics nomogram that integrates CESM-derived radiomics features and clinical parameters showed potential feasibility for predicting NAC-insensitive breast cancers.