摘要
肿瘤是全球范围内重大公共卫生问题,尽管现代医学取得了一定进展,但在精准诊疗、风险评估、药物研发等方面仍存在许多未被满足的临床需求。随着大数据技术的发展,肿瘤大数据逐渐成为推动肿瘤领域创新的重要力量。本综述通过描述肿瘤大数据的特征,包括多组学数据(如基因组、转录组、蛋白组、代谢组等)和多模态数据(如临床、影像、病理等),探讨肿瘤大数据与人工智能结合在肿瘤诊疗中的应用现状。同时,分析现阶段面临的主要挑战,如数据标准化程度不足、多模态数据融合的技术瓶颈、模型可解释性有限以及临床验证的缺乏等问题,并初步探讨可能的解决方向。最后,展望未来发展趋势,提出在肿瘤诊疗中构建专用大模型、实现多维数据融合以及设计个体化需求驱动的精准管理策略的重要性。
Tumor remains a significant global public health challenge.Despite advancements in modern medicine,there are still numerous unmet clinical needs in areas such as precision diagnosis and treatment,risk assessment,and drug development.With the evolution of big data technologies,tumor big data has gradually emerged as a crucial force driving innovation in the field of oncology.This review outlines the characteristics of tumor big data,including multi-omics data(such as genomics,transcriptomics,proteomics,and metabolomics)and multi-modal data(such as clinical,imaging,and pathological data),and explores the current applications of integrating tumor big data with artificial intelligence in tumor diagnosis and treatment.Additionally,we analyze the main challenges faced at this stage,such as insufficient data standardization,technical bottlenecks in multi-modal data integration,limited model interpretability,and a lack of clinical validation,while also discussing potential solutions.Finally,we look ahead to future trends,emphasizing the importance of constructing specialized large models for tumor diagnosis and treatment,achieving multi-dimensional data fusion,and designing personalized,demanddriven precision management strategies.
作者
肖文铉
赵珅
江一舟
Wenxuan Xiao;Shen Zhao;Yizhou Jiang(Department of Breast Surgery,Fudan University Shanghai Cancer Center,Department of Oncology,Shanghai Medical College,Fudan University,Shanghai 200032,China)
出处
《中国科学基金》
北大核心
2025年第1期153-161,共9页
Bulletin of National Natural Science Foundation of China
基金
国家自然科学基金项目(82272822)的资助。
关键词
肿瘤诊疗
肿瘤大数据
人工智能
多组学数据
多模态数据
精准医疗
tumor diagnosis and treatment
tumor big data
artificial intelligence
multi-omics data
multimodal data
precision medicine
