In the application of auxiliary clinical diagnosis and remote medical treatment, there exists a significant demand to accurately and quickly retrieve the massive medical image which are captured by different devices, correspond to different human tissues and organs, and represent a variety of diseases in the regional medical image storage and transmission system. In this paper, a variety of image retrieval modes based on regional PACS are designed, and the key technologies such as semantic annotation, visual feature library construction, etc. are analyzed to implement multi mode retrieval.%为使区域医学影像存储与传输系统(PACS)中来源于不同设备、对应不同人体组织器官和多种病症的海量医学图像资源,在辅助临床诊断和远程医疗等应用中被准确快速的检索定位,设计了基于区域PACS系统的医学图像多样化检索模式,并分析了实现多模式检索涉及的语义标注、视觉特征库构建等关键技术。
展开▼