AUTHOR=Li Yushan , Ye Maodong , Jia Baolong , Chen Linwei , Zhou Zubang TITLE=Practice of the new supervised machine learning predictive analytics for glioma patient survival after tumor resection: Experiences in a high-volume Chinese center JOURNAL=Frontiers in Surgery VOLUME=Volume 9 - 2022 YEAR=2023 URL=https://www.frontiersin.org/journals/surgery/articles/10.3389/fsurg.2022.975022 DOI=10.3389/fsurg.2022.975022 ISSN=2296-875X ABSTRACT=OBJECTIVE: To assess the effectiveness of Gradient Boosting (GB) algorithm on glioma prognosis prediction, and to explore new predictive models for glioma patient survival after tumor resection. METHODS: A cohort of 776 glioma cases (WHO II-IV) between 2010 and 2017, were obtained. Clinical characteristics and biomarker information were reviewed. Subsequently we constructed the conventional Cox survival model and three different supervised machine learning (ML) models, including random survival forest (RSF), Tree GB and Component GB model. Then the model performance was compared with each other. Lastly, we also assessed the feature importance of models. RESULTS: The concordance index of the conventional survival model, RSF, Tree GB and Component GB was 0.755, 0.830, 0.837 and 0.840 respectively. All the area under the cumulative ROCs (AUC) of both two GB models were above 0.800 at different survival time. Their calibration curves showed good calibration of the survival prediction. Meanwhile, the analysis of feature importance revealed KPS, age, tumor subtype, extent of resection and so on as crucial predictive factors. CONCLUSION: Gradient Boosting models performed better on predicting glioma patient survival after tumor resection than other models.