摘要
本文重点介绍了人工智能技术在脑肿瘤影像诊断和评估中的研究进展,并展望人工智能在脑肿瘤精准化、个性化诊治的推动作用。本研究检索日期从2015年1月1日至2019年9月15日,采用关键词检索方式,检索了PubMed和知网两个数据库,同时手工检索纳入研究的参考文献。总共纳入40项研究,其中14项研究涉及脑肿瘤诊断,6项研究涉及脑肿瘤分级,6项研究涉及预测预后,3项研究涉及评估治疗反应,11项研究涉及评估生物学特性。人工智能技术在脑肿瘤领域,如肿瘤诊断、肿瘤分级、预后预测、治疗反应和生物学特性等诸多方面均取得不错的成绩,在未来的脑肿瘤诊疗中有广阔的应用前景。然而,人工智能技术仍然存在艰巨的障碍和陷阱,如机构间数据难以聚合,犹如“数据孤岛”。面对数据共享障碍,联邦学习有望解决数据隐私问题,需要相关学者进一步探索。
This paper focuses on the research progress of artificial intelligence(AI)technology in the diagnosis and evaluation of brain tumor imaging,and looks forward to the role of AI in accurate and personalized diagnosis and treatment of brain tumors.The retrieval date of this study was from January 1,2015 to September 15,2019.Two databases,PubMed and CNKI,were searched by keyword retrieval and the references included in the study were searched manually.A total of forty studies were included in this study,of which fourteen studies involving brain tumor diagnosis,six studies involving brain tumor grading,six studies involving prognosis,three studies involving treatment response assessment,and eleven studies involving biological characteristics assessment.AI technology has made significant progress in the field of brain tumors,such as tumordiagnosis,tumorgrading,prognosisprediction,assessment of treatment response and biological characteristics.We expect that AI technology has a broad application prospect in the future diagnosis and treatment of brain tumors.However,there are still many formidable obstacles and pitfalls in AI technology such as the difficulty of aggregated data between organizations,leading to“data-island”.Facing the obstacle of data sharing,federal learning is expected to solve the data privacy problem,which needs further exploration by relevant scholars.
作者
吴双
陈蔓
周育夫
WU Shuang;CHEN Man;ZHOU Yufu(Department of Tumor Radiotherapy,The First Affiliated Hospital of Bengbu Medical College,Bengbu Anhui 233004,China)
出处
《中国医疗设备》
2020年第12期171-176,187,共7页
China Medical Devices
关键词
人工智能
机器学习
深度学习
脑肿瘤
磁共振成像
artificial intelligence
machine learning
deep learning
brain tumors
magnetic resonance imaging
