Computed tomography imaging phenotypes of hepatoblastoma identified from radiomics signatures are associated with the efficacy of neoadjuvant chemotherapy

• Published in: Pediatric radiology 2024-01, Vol.54 (1), p.58-67
• Main Authors: Chen, Yingqian, Froelich, Matthias F., Tharmaseelan, Hishan, Jiang, Hong, Wang, Yuanqi, Li, Haitao, Tao, Mingyao, Gao, Ying, Wang, Jifei, Liu, Juncheng, Schoenberg, Stefan O., Feng, Shiting, Weis, Meike
• Format: Article
• Language: English
• Subjects: Chemotherapy; Child; Cluster analysis; Clustering; Computed tomography; Embedding; Feasibility studies; Health services; Hepatoblastoma - diagnostic imaging; Hepatoblastoma - drug therapy; Humans; Imaging; Learning algorithms; Liver Neoplasms - diagnostic imaging; Liver Neoplasms - drug therapy; Machine learning; Medical imaging; Medicine; Medicine & Public Health; Neoadjuvant Therapy; Neuroradiology; Nuclear Medicine; Oncology; Original; Original Article; Pediatrics; Phenotype; Phenotypes; Qualitative analysis; Radiology; Radiomics; Retrospective Studies; Tomography, X-Ray Computed; Ultrasound; Unsupervised learning; Vector quantization
• ISSN: 1432-1998; 0301-0449; 1432-1998
• DOI: 10.1007/s00247-023-05793-5

Background Though neoadjuvant chemotherapy has been widely used in the treatment of hepatoblastoma, there still lacks an effective way to predict its effect. Objective To characterize hepatoblastoma based on radiomics image features and identify radiomics-based lesion phenotypes by unsupervised machine learning, intended to build a classifier to predict the response to neoadjuvant chemotherapy. Materials and methods In this retrospective study, we segmented the arterial phase images of 137 cases of pediatric hepatoblastoma and extracted the radiomics features using PyRadiomics. Then unsupervised k-means clustering was applied to cluster the tumors, whose result was verified by t-distributed stochastic neighbor embedding (t-SNE). The least absolute shrinkage and selection operator (LASSO) regression was used for feature selection, and the clusters were visually analyzed by radiologists. The correlations between the clusters, clinical and pathological parameters, and qualitative radiological features were analyzed. Results Hepatoblastoma was clustered into three phenotypes (homogenous type, heterogenous type, and nodulated type) based on radiomics features. The clustering results had a high correlation with response to neoadjuvant chemotherapy ( P =0.02). The epithelial ratio and cystic components in radiological features were also associated with the clusters ( P =0.029 and 0.008, respectively). Conclusions This radiomics-based cluster system may have the potential to facilitate the precise treatment of hepatoblastoma. In addition, this study further demonstrated the feasibility of using unsupervised machine learning in a disease without a proper imaging classification system. Graphical abstract