摘要
脑膜瘤是最常见的原发性中枢神经系统肿瘤之一,不同级别及亚型脑膜瘤的治疗方式存在差异,因此,脑膜瘤的早期诊断、分级及分型对全面、个体化诊疗方案的制订至关重要。影像组学和深度学习(deep learning, DL)是如今热门的研究方法,在脑膜瘤分级、鉴别诊断中应用成熟,关于预测脑膜瘤分型、基因表型等方面的研究日益增多。二者具有快速精准、全自动化学习、无创性及客观性等特点,能为疾病提供更加准确的诊断、治疗和预后预测,为临床治疗及患者预后提供重要帮助。本文将从脑膜瘤术前分级分型、鉴别诊断、预后复发及基因表型预测几个方面来总结并分析影像组学及DL在脑膜瘤中的研究进展,对现有研究的成果、局限性以及未来的改进措施和发展方向进行总结,以期推进影像组学与DL在脑膜瘤诊疗中的应用进程。
Meningioma is one of the most common primary central nervous system tumours,treatment modalities vary between different grades and subtypes of meningioma,so early diagnosis,grading,and typing of meningiomas are critical to the development of a comprehensive and individualized treatment plan.Radiomics and deep learning(DL)are now popular research methods,and are well established for the classification and differential diagnosis of meningiomas.Both have the characteristics of fast and accurate,fully automated learning,non-invasive and objective,which can provide more accurate diagnosis,treatment and prognosis prediction of the disease.In this paper,we will summarise and analyse the research progress of imaging histology and DL in meningioma in terms of preoperative grading and staging,differential diagnosis,prognosis of recurrence and prediction of genetic phenotype,and summarise the achievements and limitations of existing studies as well as future improvement measures and development directions,to promote the application of imaging omics and DL in the diagnosis and treatment of meningioma.
作者
杨慧敏
李文鑫
刘艳美
禹雯婧
王倩倩
姜兴岳
刘新疆
YANG Huimin;LI Wenxin;LIU Yanmei;YU Wenjing;WANG Qianqian;JIANG Xingyue;LIU Xinjiang(Department of Radiology,Affiliated Hospital of Binzhou Medical College,Binzhou 256603,China;Department of Radiology,Shanghai Pudong Hospital(Pudong Hospital of Hehai University),Shanghai 201399,China)
出处
《磁共振成像》
CAS
CSCD
北大核心
2023年第6期124-128,共5页
Chinese Journal of Magnetic Resonance Imaging
基金
上海市卫生健康委员会卫生行业临床研究专项立项项目(编号:202140266)
浦东新区卫生健康委卫生科技项目(编号:PW2020A-35)。
关键词
脑膜瘤
深度学习
影像组学
分级分型
鉴别诊断
预后预测
磁共振成像
meningioma
deep learning
radiomics
grading
differential diagnosis
prognostic prediction
magnetic resonance imaging
