文档介绍：Optimized GPU Framework for Ultrasound Color
Flow Imaging
Zhengjuan Fan Dan Shi, Dong
Computer Science puter Science College
Sichuan University Sichuan University
Chengdu, China Chengdu, China
cuiwhere@ standing06@, dongcliu@
Abstract— A GPU framework for ultrasound color flow imaging the use of GPU for the signal-processing algorithm and focuses
(CFI) based on auto-correlation is presented. The parallel CFI on color Doppler signal processing [4]. Whereas a kind of
processing framework implementation is mainly based on CUDA flexible, robust, applied GPU-based framework for CFI still
performance features, such as the memory selection strategy, need to be optimized. Parallel algorithms for the framework are
applicable thread structure and high-throughput bandwidth. required to further research by deeply mining GPU
Parallel convolution algorithm and multi-channel championship performance characteristics and by seeking for some efficient
algorithm are proposed. This CFI method achieves a frame rate general GPU-based parallel algorithms for signal processing.
of 300 fps from the Doppler signal, in which the number of scan
lines is 44, the number of samples along the axial line is 510 and The main purpose of this study focuses on the design and
ensemble size is 16. implementation of parallel algorithms of autocorrelation-based
ultra