人工智能在前列腺癌MRI诊断中的研究进展  

Research progress of artificial intelligence in MRI diagnosis of prostate cancer

作　　者：黄兴鸿 王伟[1] 曹学香 丁偕[2] 王培军[1] HUANG Xing-hong;WANG Wei;CAO Xue-xiang;DING Xie;WANG Pei-jun(Dept.of Radiology,Tongji Hospital,School of Medicine,Tongji University,Shanghai 200065,China;Wonders Information Co.,Ltd,Shanghai 201112,China)

机构地区：[1]同济大学附属同济医院医学影像科,上海200065 [2]万达信息股份有限公司,上海201112

出　　处：《同济大学学报（医学版）》2021年第4期562-567,共6页Journal of Tongji University(Medical Science)

基　　金：上海市科学技术委员会“科技创新行动计划”项目(17411952300);上海市同济医院院级青年科研项目(TJ1912)。

摘　　要：人工智能(artificial intelligence,AI)近年来在医疗领域得到了快速发展,本文回顾了AI技术在前列腺磁共振成像(magnetic resonance imaging,MRI)领域的发展情况。本文介绍AI相关研究方法,包括机器学习及其常见分类,分析其在前列腺MRI研究中的应用方向。然后综合介绍目前AI技术在前列腺方向的检测、分类、分级以及诊疗、监控等方面的应用现状。虽然目前的研究已经取得了很多令人满意的结果,但是在更加广泛的临床应用之前,还需要进行更多的可靠性验证。未来AI的算法、软件可能在前列腺肿瘤的诊疗领域辅助临床医生进行决策。In recent years,artificial intelligence(AI)has developed rapidly in the medical fields.This article reviews the development of AI in magnetic resonance imaging(MRI)diagnosis of prostate cancer.The article first introduce the research methods of artificial intelligence,including machine learning and its common classifications,and the research direction in prostate MRI.Then it introduces the application of artificial intelligence in the detection,classification,grading,diagnosis and treatment of prostate cancer.Although satisfactory research results have been achieved currently,which need to be further verified for its reliability in extensive clinical application.Hopefully in the future the algorithms and software of artificial intelligence can assist clinicians to make decisions for diagnosis and treatment of prostate cancer.

关 键 词：自动分类 前列腺癌 磁共振成像 多参数 神经网络 

分 类 号：R445.2[医药卫生—影像医学与核医学]