高强度聚焦超声(HIFU)治疗系统中B超监控图像由于存在伪像和噪声而很难实现图像配准,给靶区定位带来了困难,基于此,提出利用基于超声散斑的相位相关对B超图像进行配准并验证其可行性。首先,根据超声散斑原理和散斑伪随机性,在Matlab7.0软件环境下,利用基于超声散斑的相位相关对从B超图像中提取出散斑图像进行配准,根据得出的位移值配准对应的B超图像;其次,为了与基于超声散斑的图像配准进行对比,对B超图像进行直接相位相关配准;最后,观察配准结果并进行相似度测定。结果显示,基于超声散斑的相位相关算法能够实现对存在伪像和噪声的B超图像的配准,具有较高的精确性和鲁棒性,优于对B超图像直接相位相关算法;同时也验证了基于超声散斑配准B超图像的可行性。
The existence of phantoms and other ultrasound noises in Ultrasound-Guided HIFU systems' B-mode ultrasound images will hamper the B-mode ultrasound image registration and the locating of the target.In order to solve this problem,this paper proposes a new method of B-mode ultrasound image registration based on the phase correlation of ultrasonic speckle.Firstly,under the principles and the pseudorandom feature of the speckle,the speckle images and registration,the B-mode ultrasound images are picked up by using the phase correlation of ultrasonic speckle through the moving values of images,on the platform of Maltab 7.0.Secondly,the direct registration of the phase correlation is compared with the registration of the speckle in B-mode ultrasound images.Thirdly,the results of registration are analyzed and the measurement is made of the similarity degree.It is shown that the phase correlation of the speckle could achieve the image registration among the B-mode ultrasound images with phantoms and noises.This registration method enjoys a high robustness and accuracy,and is better than the direct registration of phase correlation.The feasibility of the registration based on speckle is demonstrated by experiments.