多模态医学影像融合识别技术研究进展  

Research Progress of Multi-Model Medical Image Fusion and Recognition

作　　者：周涛[1] 陆惠玲[1] 陈志强[2] 马竟先[1] 

机构地区：[1]宁夏医科大学理学院,银川750004 [2]宁夏医科大学附属医院放射科,银川750004

出　　处：《生物医学工程学杂志》2013年第5期1117-1122,共6页Journal of Biomedical Engineering

基　　金：国家自然科学基金资助项目(81160183);教育部"春晖计划"资助项目(Z2011051);宁夏自然科学基金资助项目(NZ12179);宁夏卫生厅重点科研基金资助项目(2011033);宁夏高等学校科研重点基金资助项目(NGY2011042);宁夏医科大学特殊人才基金资助项目(XT2011004);宁夏医科大学青年基金基金资助项目(XQ2011011)

摘　　要：医学影像融合识别在病灶定位、癌症分期、治疗效果评估等方面有着广泛的应用。本文对多模态医学影像融合技术的研究进行了分析和总结,首先分析多模态医学影像融合识别问题,着重讨论了多模态医学影像融合识别的优势、关键步骤等;然后从算法的角度综述了三种融合识别策略,并讨论了四种融合识别结构,最后对目前所面临的问题和进一步的研究方向进行了探讨。Medical image fusion and recognition has a wide range of applications, such as focal location, cancer stag- ing and treatment effect assessment. Multi-model medical image fusion and recognition are analyzed and summarized in this paper. Firstly, the question of multi-model medical image fusion and recognition is discussed, and its advan- tage and key steps are discussed. Secondly, three fusion strategies are reviewed from the point of algorithm, and four fusion recognition structures are discussed. Thirdly, difficulties, challenges and possible future research direction are discussed.

关 键 词：多模态 医学影像 融合识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]