Study On Ultrasonic Extended-Field-of-View Imaging Method

As a novel ultrasonic imaging technology, ultrasonic extended-field-of-view (EFOV) imaging overcomes the drawback of traditional ultrasonic imaging which can only generate ultrasonic images with very limited vision field, and is widely used in medical institution in recent years.The theory and realization methods of ultrasonic EFOV imaging technology are studied, and some key technologies involved are analyzed and discussed. Improved algorithms for ultrasonic speckle noise suppression and image registration are proposed in this paper. By analyzing the statistic model of the speckle in ultrasonic images and studying the influences of log compression to speckle statistic, a robust speckle noise suppression algorithm is proposed, which estimates the dynamic range parameter of the log suppression model using an iterative method and overcomes the shortage of the current methods which can not precisely estimate the dynamic range parameter. Experiment results show that the proposed method can restrain the speckle noise and improve the precision of image registration. For ultrasonic images registration problem, most of current ultrasonic EFOV technologies only use the translation and rotation model in image registration, and the speed and accuracy of the registration algorithms are not satisfactory. In this paper, a coarse to fine registration algorithm is proposed, which firstly estimates the approximate translation by coarse image registration based on block matching method. Then the affine transform is used as the transform model between image frames and Levenberg-Maquardt method is used to search the optimal affine transform parameters. The effectiveness of the proposed algorithm is demonstrated by experiment results.