前列腺癌风险预测模型研究与应用分析——以数据驱动的视角  

Research and Application Analysis of Risk Prediction Models for Prostate Cancer from a Data-driven Perspective

作　　者：吴杰[1,2] 陈亚兰[3] 魏雪栋[2] 黄玉华[2] 侯建全[1,2] 林宇鑫[2] WU Jie;CHEN Yalan;WEI Xuedong;HUANG Yuhua;HOU Jianquan;LIN Yuxin(Department of Urology,Dushu Lake Hospital Affiliated to Soochow University,Suzhou 215000,China;Department of Urology,The First Affiliated Hospital of Soochow University,Suzhou 215006,China;Department of Medical Informatics,School of Medicine,Nantong University,Nantong 226001,China)

机构地区：[1]苏州大学附属独墅湖医院泌尿外科,苏州215000 [2]苏州大学附属第一医院泌尿外科,苏州215006 [3]南通大学医学院医学信息学系,南通226001

出　　处：《医学信息学杂志》2023年第12期40-46,共7页Journal of Medical Informatics

基　　金：国家自然科学基金青年科学基金项目(项目编号:32200533);国家自然科学基金青年科学基金项目(项目编号:82102186);江苏省重点研发计划(社会发展)项目-重点项目-临床前沿技术(项目编号:BE2020655)。

摘　　要：目的/意义回顾性分析前列腺癌风险预测模型的研究进展,为前列腺癌智能化诊疗体系建设提供参考。方法/过程通过文献挖掘分析,探讨分子、影像、个体、群体等不同组学层次评价指标对前列腺癌的诊断意义,比较不同计算模型的关键特征变量和临床应用价值。结果/结论现有模型具有计算简便、可行性强等优势,但也存在预测精度有限、泛化能力不足等局限。整合多组学数据和人工智能模型对于前列腺癌的医学信息学研究和智慧医疗体系建设具有重要意义。Purpose/Significance The paper retrospectively analyzes the research progress on risk prediction models for prostate cancer,and provides references for the construction of prostate cancer intelligent diagnosis and treatment system.Method/Process Through literature mining and analysis,the significance of molecular,imaging,individual,population and other omics level evaluation indexes in the diagnosis of prostate cancer is discussed,and the key characteristic variables and clinical application value of different calculation models are compared.Result/Conclusion The existing models have advantages of easy calculation and strong feasibility,but they also have limitations such as limited prediction accuracy and insufficient generalization ability.The integration of multi-omics data and artificial intelligence models is of great significance for medical informatics research and smart medical system construction of prostate cancer.

关 键 词：前列腺癌 预测变量 计算模型 智慧医疗 

分 类 号：R-058[医药卫生]