Hongpeng Chu1, Zelong Liu2, Wen Liang3, Qian Zhou4, Ying Zhang1, Kai Lei1, Mimi Tang5, Yiheng Cao4, Shuling Chen2, Sui Peng6,7, Ming Kuang8,9. 1. Department of Liver Surgery, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. 2. Department of Medical Ultrasonics, Institute of Diagnostic and Interventional Ultrasound, The First Affiliated Hospital of Sun Yat-sen University, 58 Zhong Shan Road 2, Guangzhou, 510080, China. 3. Department of Radiology, The Zhu Jiang Hospital of Southern Medical University, Guangzhou, China. 4. Clinical Trial Unit, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. 5. Department of Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. 6. Clinical Trial Unit, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. pengsui@vip.163.com. 7. Department of Gastroenterology and Hepatology, The First Affiliated Hospital of Sun Yat-sen University, 58 Zhong Shan Road 2, Guangzhou, 510080, China. pengsui@vip.163.com. 8. Department of Liver Surgery, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China. kuangminda@hotmail.com. 9. Department of Medical Ultrasonics, Institute of Diagnostic and Interventional Ultrasound, The First Affiliated Hospital of Sun Yat-sen University, 58 Zhong Shan Road 2, Guangzhou, 510080, China. kuangminda@hotmail.com.
Abstract
OBJECTIVES: To investigate and compare radiomics and clinical information for preoperative prediction of futile resection in intrahepatic cholangiocarcinoma (ICC). METHODS: A total of 203 ICC patients from two centers were included and randomly allocated with a ratio of 7:3 into the training cohort and the validation cohort. Clinical characteristics and radiomics features were selected using random forest algorithm and logistic models to construct a clinical model and a radiomics model, respectively. A combined logistic model that incorporated the developed radiomics signature and clinical risk factors was then built. The performance of these models was evaluated and compared by plotting the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC). RESULTS: The radiomics model showed a higher AUC than the clinical model in the validation cohort (AUC: 0.804 (95% CI: 0.697, 0.912) vs. 0.590 (95% CI: 0.415, 0.765), p = 0.043) for predicting futile resection in ICC. The radiomics model reached a sensitivity of 0.846 (95% CI: 0.546, 0.981) and a specificity of 0.771 (95% CI: 0.627, 0.880) in the validation cohort. Moreover, the radiomics model had comparable AUCs with the combined model in training and validation cohorts. CONCLUSIONS: We presented an internally validated radiomics model for the prediction of futile resection in ICC patients. Compared with clinical information, radiomics using CT images had greater potential for predicting futile resection accurately before surgery. KEY POINTS: • Radiomics model using CT images could predict futile resection in intrahepatic cholangiocarcinoma preoperatively. • Radiomics model using CT images was superior to clinical information for predicting futile resection accurately before surgery.
OBJECTIVES: To investigate and compare radiomics and clinical information for preoperative prediction of futile resection in intrahepatic cholangiocarcinoma (ICC). METHODS: A total of 203 ICC patients from two centers were included and randomly allocated with a ratio of 7:3 into the training cohort and the validation cohort. Clinical characteristics and radiomics features were selected using random forest algorithm and logistic models to construct a clinical model and a radiomics model, respectively. A combined logistic model that incorporated the developed radiomics signature and clinical risk factors was then built. The performance of these models was evaluated and compared by plotting the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC). RESULTS: The radiomics model showed a higher AUC than the clinical model in the validation cohort (AUC: 0.804 (95% CI: 0.697, 0.912) vs. 0.590 (95% CI: 0.415, 0.765), p = 0.043) for predicting futile resection in ICC. The radiomics model reached a sensitivity of 0.846 (95% CI: 0.546, 0.981) and a specificity of 0.771 (95% CI: 0.627, 0.880) in the validation cohort. Moreover, the radiomics model had comparable AUCs with the combined model in training and validation cohorts. CONCLUSIONS: We presented an internally validated radiomics model for the prediction of futile resection in ICC patients. Compared with clinical information, radiomics using CT images had greater potential for predicting futile resection accurately before surgery. KEY POINTS: • Radiomics model using CT images could predict futile resection in intrahepatic cholangiocarcinoma preoperatively. • Radiomics model using CT images was superior to clinical information for predicting futile resection accurately before surgery.
Authors: Wenya Linda Bi; Ahmed Hosny; Matthew B Schabath; Maryellen L Giger; Nicolai J Birkbak; Alireza Mehrtash; Tavis Allison; Omar Arnaout; Christopher Abbosh; Ian F Dunn; Raymond H Mak; Rulla M Tamimi; Clare M Tempany; Charles Swanton; Udo Hoffmann; Lawrence H Schwartz; Robert J Gillies; Raymond Y Huang; Hugo J W L Aerts Journal: CA Cancer J Clin Date: 2019-02-05 Impact factor: 508.702
Authors: Lukas Müller; Aline Mähringer-Kunz; Simon Johannes Gairing; Friedrich Foerster; Arndt Weinmann; Fabian Bartsch; Lisa-Katharina Heuft; Janine Baumgart; Christoph Düber; Felix Hahn; Roman Kloeckner Journal: J Clin Med Date: 2021-05-12 Impact factor: 4.241