Communication optimization and auto load balancing in parallel OSEM algorithm for fully 3-D SPECT reconstruction

In order to improve the computation speed of ordered subset expectation maximization (OSEM) algorithm for fully 3-D single photon emission computed tomography (SPECT) reconstruction, a parallelizing scheme of OSEM reconstruction algorithm was implemented on an experimental beowulf-type cluster and impact factors on the parallel efficiency were investigated. Two approaches were employed to improve the efficiency: (1) the communication cost was minimized via overlapping communication with computation and (2) the idle time of processes was reduced by auto load balancing. Performance of the optimized parallel algorithm was evaluated in terms of computation time, speedup factor and parallel efficiency. Improvements were observed after optimization. The efficiency was raised from 83.86% to 92.07% in fully 3-D 128 times 128 times 128 SPECT reconstruction

[1]  Rajeev Thakur,et al.  Improving the Performance of Collective Operations in MPICH , 2003, PVM/MPI.

[2]  L. M. Patnaik,et al.  High-speed computation of the EM algorithm for PET image reconstruction , 1994 .

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

[4]  Freek J. Beekman,et al.  Efficient fully 3-D iterative SPECT reconstruction with Monte Carlo-based scatter compensation , 2002, IEEE Transactions on Medical Imaging.

[5]  M. Casey,et al.  SPMD cluster-based parallel 3D OSEM , 2002, 2002 IEEE Nuclear Science Symposium Conference Record.

[6]  Michael E. Casey,et al.  HeinzelCluster: accelerated reconstruction for FORE and OSEM3D. , 2002 .

[7]  A. Celler,et al.  Implementation of an analytically based scatter correction in SPECT reconstructions , 2003, 2003 IEEE Nuclear Science Symposium. Conference Record (IEEE Cat. No.03CH37515).

[8]  M.D. Jones,et al.  Parallel programming for OSEM reconstruction with MPI, OpenMP, and hybrid MPI-OpenMP , 2004, IEEE Symposium Conference Record Nuclear Science 2004..

[9]  Impact of attenuation, scatter and depth dependent collimator response correction in 3-D cardiac SPECT imaging: a Monte Carlo investigation , 2004, IEEE Symposium Conference Record Nuclear Science 2004..

[10]  D. Scott Wills,et al.  Parallelized formulation of the maximum likelihood-expectation maximization algorithm for fine-grain message-passing architectures , 1995, IEEE Trans. Medical Imaging.