基于多随机经验核的弥漫大B细胞淋巴瘤复发预测  

Precise Prediction of Diffuse Large B-Cell Lymphoma based on Multiple Random Empirical Kernel Learning Machine

作　　者：李雪玲 赵艳琳 张岩波[1,2,3] 余红梅 周洁[4] 李琼 王俊霞[1,2] 乔宇 张高源 赵志强 罗艳虹 Li Xueling;Zhan Yanlin;Zhang Yanbo(Department of Health Statistics,School of Public Health,Shanxi Medical University(030001),Taiyuan)

机构地区：[1]山西医科大学公共卫生学院卫生统计教研室,030001 [2]重大疾病风险评估山西省重点实验室 [3]煤炭环境致病与防治教育部重点实验室 [4]山西省肿瘤医院核医学PET/CT中心 [5]山西省肿瘤医院血液科

出　　处：《中国卫生统计》2024年第3期339-343,共5页Chinese Journal of Health Statistics

基　　金：山西省科技厅应用基础研究计划面上项目(202103021224245);国家自然科学基金青年科学基金(81502897,82273742);山西医科大学博士启动基金(BS2017029)。

摘　　要：目的基于多随机经验核分类器构建弥漫大B细胞淋巴瘤完全缓解后两年内复发情况的预测模型,为患者的治疗提供决策依据。方法利用山西省某三甲医院2010-2020年电子病历库中符合本研究要求的445名患者信息,基于五种常见类别不平衡处理方法以及多随机经验核分类器构建复发预测模型,并与五种分类器进行比较。结果基于SMOTE Tomek Links+多随机经验核分类器的复发预测模型取得了最优的分类性能(accuracy=0.89,precision=0.87,recall=0.92,f1-Score=0.89,brier score=0.11)。结论对DLBCL实际数据集,本文使用SMOTE Tomek links处理不平衡数据并构建多随机经验核模型,模型性能达到最优的同时计算复杂度也不高,可为DLBCL复发预测提供有力参考。Objectives To construct a prediction model of relapse in diffuse large B-cell lymphoma within two years after complete remission based on multiple randomized empirical kernel learning machine to provide a basis for patient treatment decisions.Methods Using the information of 445 patients who met the requirements of this study in the electronic medical record database of a tertiary hospital in Shanxi Province from 2010 to 2020,a relapse prediction model was constructed based on five common categories of imbalance treatment methods and a multiple stochastic empirical kernel learning machine,and compared with the five classifiers.Results The recurrence prediction model based on SMOTE Tomek Links+multiple randomized empirical kernel learning machine achieved optimal classification performance(accuracy=0.89,precision=0.87,recall=0.92,f1-Score=0.89,brier score=0.11).Conclusion For the actual DLBCL dataset,in this paper,we used SMOTE Tomek links to process the imbalance data and construct a multiple randomized empirical kernel learning machine,which achieves the optimal model performance with low computational complexity and can provide a powerful reference for DLBCL recurrence prediction.

关 键 词：弥漫大B细胞淋巴瘤 复发预测 经验核映射 类别不平衡 

分 类 号：R195.1[医药卫生—卫生统计学]