Radiomics models to predict bone marrow metastasis of neuroblastoma using CT

作     者：Xiong Chen Qinchang Chen Yuanfang Liu Ya Qiu Lin Lv Zhengtao Zhang Xuntao Yin Fangpeng Shu 

作者机构：Department of Paediatric UrologyGuangzhou Women and Children's Medical CenterGuangzhou Medical UniversityGuangzhouChina Department of Paediatric SurgeryGuangzhou Institute of PaediatricsGuangzhou Medical UniversityGuangzhouChina Department of Pediatric CardiologyGuangdong Provincial People's HospitalGuangdong Academy of Medical SciencesGuangdong Cardiovascular InstituteGuangdong Provincial Key Laboratory of Structural Heart DiseaseGuangzhouChina Department of RadiologySun Yat‐sen Memorial HospitalSun Yat‐sen UniversityGuangzhouChina Department of Radiologythe First People's Hospital of Kashi PrefectureKashiChina Medical SchoolSun Yat‐sen UniversityGuangzhouChina Department of RadiologyGuangzhou Women and Children's Medical CenterGuangzhouChina 

出 版 物：《Cancer Innovation》 (肿瘤学创新（英文）)

年 卷 期：2024年第3卷第5期

页      面：71-83页

学科分类：1002[医学-临床医学] 100214[医学-肿瘤学] 10[医学] 

基　　金：Science and Technology Projects in Guangzhou,Grant/Award Number:202201011843 National Natural Science Foundation of China,Grant/Award Numbers:82103544,82202251 

主　　题：bone marrow metastasis machine learning neuroblastoma radiomics 

摘      要：Background:Bone marrow is the leading site for metastasis from neuroblastoma and affects the prognosis of patients with ***,the accurate diagnosis of bone marrow metastasis is limited by the high spatial and temporal heterogeneity of *** analysis has been applied in various cancers to build accurate diagnostic models but has not yet been applied to bone marrow metastasis of ***:We retrospectively collected information from 187 patients pathologically diagnosed with neuroblastoma and divided them into training and validation sets in a ratio of 7:3.A total of 2632 radiomics features were retrieved from venous and arterial phases of contrastenhanced computed tomography(CT),and nine machine learning approaches were used to build radiomics models,including multilayer perceptron(MLP),extreme gradient boosting,and random *** also constructed radiomics‐clinical models that combined radiomics features with clinical predictors such as age,gender,ascites,and lymph gland *** performance of the models was evaluated with receiver operating characteristics(ROC)curves,calibration curves,and risk decile ***:The MLP radiomics model yielded an area under the ROC curve(AUC)of 0.97(95%confidence interval[CI]:0.95–0.99)on the training set and 0.90(95%CI:0.82–0.95)on the validation *** radiomics‐clinical model using an MLP yielded an AUC of 0.93(95%CI:0.89–0.96)on the training set and 0.91(95%CI:0.85–0.97)on the validation ***:MLP‐based radiomics and radiomics‐clinical models can precisely predict bone marrow metastasis in patients with neuroblastoma.