Objective To investigate the value of enhanced chest CT radiomics features in predicting the efficacy of epirubicin combined with ifosfamide in patients with pulmonary metastases from soft tissue sarcoma Methods In a retrospective analysis of 51 patients with pulmonary metastases from soft tissue sarcoma, all received the chemotherapy regimen of epirubicin combined with ifosfamide, and the efficacy was evaluated by the RECIST(1.1). Each patient selected 1 or 2 chemotherapy regimens. Lung metastases were used as target lesions (86 target lesions total), and the patients were divided into a progression group (n = 29) and a nonprogressive group (n = 57). The nonprogressive group included a stable group (n = 34) and a partial response group (n = 23). Information of lung metastases was extracted from enhanced CT images before chemotherapy, and all lesions were delineated by ITK-SNAP software manually or semiautomatically. The decision tree classifier had a better effect in all radiomics models. The receiver operating characteristic (ROC) curve was plotted to evaluate the predictive performance of the model for progression vs. nonprogression. Results In total, 851 enhanced CT radiomics features were extracted for each target lesion and finally reduced to 2 radiomics features, which were used to construct a radiomics model. The areas under the curve (AUCs) of the model for predicting the progression of lesions were 0.917 and 0.856 in the training and test groups. Conclusion The model established based on the radiomic features of chest enhanced CT before treatment has certain predictive value for the chemotherapy efficacy of patients with soft tissue sarcoma lung metastases.

Objective: To investigate the value of enhanced chest CT radiomics features in predicting the efficacy of epirubicin combined with ifosfamide in patients with pulmonary metastases from soft tissue sarcoma Methods: In a retrospective analysis of 51 patients with pulmonary metastases from soft tissue sarcoma, all received the chemotherapy regimen of epirubicin combined with ifosfamide, and the efficacy was evaluated by the RECIST（1.1）. Each patient selected 1 or 2 chemotherapy regimens. Lung metastases were used as target lesions (86 target lesions total), and the patients were divided into a progression group (n=29) and a nonprogressive group (n=57). The nonprogressive group included a stable group (n=34) and a partial response group (n=23). Information of lung metastases was extracted from enhanced CT images before chemotherapy, and all lesions were delineated by ITK-SNAP software manually or semiautomatically. The decision tree classifier had a better effect in all radiomics models. The receiver operating characteristic (ROC) curve was plotted to evaluate the predictive performance of the model for progression vs. non-progression.Results: In total, 851 enhanced CT radiomics features were extracted for each target lesion and finally reduced to 2 radiomics features, which were used to construct a radiomics model. The areas under the curve (AUCs) of the model for predicting the progression of lesions were 0.917 and 0.856 in the training and test groups.Conclusion: The model established based on the radiomics features of chest enhanced CT before treatment has certain predictive value for the chemotherapy efficacy of patients with soft tissue sarcoma lung metastases.

Objective To investigate the value of enhanced chest CT radiomics features in predicting the efficacy of epirubicin combined with ifosfamide in patients with pulmonary metastases from soft tissue sarcoma Methods In a retrospective analysis of 51 patients with pulmonary metastases from soft tissue sarcoma, all received the chemotherapy regimen of epirubicin combined with ifosfamide, and the efficacy was evaluated by the RECIST(1.1). Each patient selected 1 or 2 chemotherapy regimens. Lung metastases were used as target lesions (86 target lesions total), and the patients were divided into a progression group (n = 29) and a nonprogressive group (n = 57). The nonprogressive group included a stable group (n = 34) and a partial response group (n = 23). Information of lung metastases was extracted from enhanced CT images before chemotherapy, and all lesions were delineated by ITK-SNAP software manually or semiautomatically. The decision tree classifier had a better effect in all radiomics models. The receiver operating characteristic (ROC) curve was plotted to evaluate the predictive performance of the model for progression vs. non-progression. Results In total, 851 enhanced CT radiomics features were extracted for each target lesion and finally reduced to 2 radiomics features, which were used to construct a radiomics model. The areas under the curve (AUCs) of the model for predicting the progression of lesions were 0.917 and 0.856 in the training and test groups. Conclusion The model established based on the radiomics features of chest enhanced CT before treatment has certain predictive value for the chemotherapy efficacy of patients with soft tissue sarcoma lung metastases.