SimDeCCom: Simultaneous De-Cluttering and Compression for Real-Time Ultrasound Imaging

Real-time diagnostic imaging is a major advantage of ultrasound medical systems. Image distortion, caused by tissue in-homogeneities and cluttering effects, may be reduced by applying aperture-domain processing. In order to accommodate real-time processing of array signals, software-based ultrasonic imaging architectures have been developed. Such systems consist of a front-end analog data acquisition component, and a back-end digital signal processor. However, the main bottleneck lies in the data transfer rate (typically Glga-bits/sec). In order to address this limitation, it was previously suggested to integrate real-time data compression into the design process of the imaging systems. We take this approach one step further, and propose a combined compression and de-cluttering scheme, namely: SimDeCCom (Simultaneous De-Cluttering and Compression for Real-Time Ultrasound Imaging). SimDeCCom is an adaptive RF data compression scheme, where compression is adaptively applied such that maximal fidelity is obtained with respect to a desired output image, which is a de-cluttered version of the original image. In addition, the proposed algorithm enables to attain a chosen target bit-rate, such as to accommodate hardware implementation constraints. We demonstrate SimDeCCom's performance for echo-cardiac data.