医学图像配准对提高临床诊断治疗水平、病情监测以及放射治疗和手术导航等具有重要的研究意义。医学图像配准技术主要包括刚体与非刚体两种类型,基于刚体的图像配准算法发展迅速已广泛应用于临床,非刚体图像配准技术尚在发展阶段。本文在陈述配准技术框架及其各模块的基础上,综述了近年来一些典型的非刚体配准算法,包括算法原理、主要特点、适用情况及其改进方向等。最后,对非刚体配准技术的发展进行了展望。
Medical image registration plays a significant role in clinical diagnosis and treatment,illness monitoring,radiotherapy,surgical navigation and so on.Medical image registration algorithms are classified into rigid registration and non-rigid registration.At present,the rigid registration algorithms have been developed and are used well in actual clinical application.However,the non-rigid medical image registration algorithms are still in developing stage.Based on the summarization of the medical image registration framework and each module of the framework,an overview of non-rigid medical image registration algorithms is presented in this paper,which includes algorithm theory,main characteristics,applicable condition and the improvement direction,etc.Finally, the trend of non-rigid registration algorithms in future is prospected.