基于CUDA的图形处理器加速锥束CT重建算法的研究

摘要

锥束CT图像重建数据量巨大、运算复杂度高，重建时间长，难以满足实际应用的需求。研究基于CUDA的图形处理器加速锥束CT重建算法的方案，通过有效的并行策略来提高滤波和反投影过程的时间，并利用常数存储器和纹理存储器来提高数据访存效率。实验证明在保证重建质量的情况下，重建速度可以提高82倍。%Cone beam CT image reconstruction has huge data volume and high operation complexity,the time of image reconstruction is too long to meet the needs of practical applications.In this paper we study the acceleration solution of cone beam CT image reconstruction algo-rithm with the CUDA-based GPU.It improves the filtering and back projection process time through effective parallel strategy,and improves data access and storage efficiency using constant memory and texture memory.Experimental results show that there can have 82 times im-provement in reconstruction speed under the condition of ensuring the quality of reconstruction.