Fast Implementation of Iterative Reconstruction with Exact Ray-Driven Projector on GPUs

Abstract Iterative methods are popular choices in image reconstruction fields due to their capability of recovering object information from incomplete acquisition data. However, the computation process involves frequent uses of forward and backward projections that are computationally expensive. Past research has proved that a forward projector that can produce high quality images is crucial to achieve a good convergence rate. In this paper a high performance iterative reconstruction framework is introduced, where two most popular iterative algorithms: Simultaneous Algebraic Reconstruction Technique (SART) and Ordered-subsets Expectation Maximization (OSEM) are supported. The framework utilizes Siddon's ray-driven method to generate forward projected images. Benefited from functionalities offered by current generation of graphics processing units (GPUs), it achieves better performance when compared to previous GPU implementations that use grid-interpolated methods, on top of the significant speedups over CPU-based solutions.

[1]  A. Kak,et al.  Simultaneous Algebraic Reconstruction Technique (SART): A Superior Implementation of the Art Algorithm , 1984, Ultrasonic imaging.

[2]  Huaxia Zhao,et al.  Fast ray-tracing technique to calculate line integral paths in voxel arrays , 2003, 2003 IEEE Nuclear Science Symposium. Conference Record (IEEE Cat. No.03CH37515).

[3]  Bruce H. Hasegawa,et al.  Stream processors: a new platform for Monte Carlo calculations , 2008 .

[4]  J. Bian,et al.  Comparison of reconstruction algorithms for digital breast tomosynthesis , 2009, 0908.2610.

[5]  James T Dobbins,et al.  Digital x-ray tomosynthesis: current state of the art and clinical potential. , 2003, Physics in medicine and biology.

[6]  Fang Xu,et al.  Accelerating popular tomographic reconstruction algorithms on commodity PC graphics hardware , 2005, IEEE Transactions on Nuclear Science.

[7]  Fang Xu,et al.  High performance tomosynthesis enabled via a GPU-based iterative reconstruction framework , 2009, Medical Imaging.

[8]  L. Feldkamp,et al.  Practical cone-beam algorithm , 1984 .

[9]  H. Malcolm Hudson,et al.  Accelerated image reconstruction using ordered subsets of projection data , 1994, IEEE Trans. Medical Imaging.

[10]  Klaus Mueller,et al.  A comparative study of popular interpolation and integration methods for use in computed tomography , 2006, 3rd IEEE International Symposium on Biomedical Imaging: Nano to Macro, 2006..

[11]  R. Siddon.  Fast calculation of the exact radiological path for a three-dimensional CT array. , 1985, Medical physics.

[12]  Zhengrong Liang,et al.  Speedup OS-EM Image Reconstruction by PC Graphics Card Technologies for Quantitative SPECT with Varying Focal-Length Fan-Beam Collimation. , 2005, IEEE transactions on nuclear science.

[13]  Bjorn De Sutter,et al.  A Fast Algorithm to Calculate the Exact Radiological Path through a Pixel or Voxel Space , 1998 .

[14]  Ignace Lemahieu,et al.  A fast, cache-aware algorithm for the calculation of radiological paths exploiting subword parallelism , 1999, J. Syst. Archit..