Research Article

Parallel computing for simultaneous iterative tomographic imaging by graphics processing units

Published: 2016-5-20

Journal: SPIE Proceedings

DOI: 10.1117/12.2223466

Abstract

In this paper, we address the problem of accelerating inversion algorithms for nonlinear acoustic tomographic imaging by parallel computing on graphics processing units (GPUs). Nonlinear inversion algorithms for tomographic imaging often rely on iterative algorithms for solving an inverse problem, thus computationally intensive. We study the simultaneous iterative reconstruction technique (SIRT) for the multiple-input-multiple-output (MIMO) tomography algorithm which enables parallel computations of the grid points as well as the parallel execution of multiple source excitation. Using graphics processing units (GPUs) and the Compute Unified Device Architecture (CUDA) programming model an overall improvement of 26.33x was achieved when combining both approaches compared with sequential algorithms. Furthermore we propose an adaptive iterative relaxation factor and the use of non-uniform weights to improve the overall convergence of the algorithm. Using these techniques, fast computations can be performed in parallel without the loss of image quality during the reconstruction process.

Faculty Members

  • Enyue Lu - Salisbury Univ. (United States)
  • Pedro D. Bello-Maldonado - Univ. of Illinois Urbana-Champaign (United States)
  • Colleen Rogers - Salisbury Univ. (United States)
  • Ricardo López - Univ. of Puerto Rico (United States)
  • Yuanwei Jin - Univ. of Maryland Eastern Shore (United States)

Themes

  • Algorithm convergence
  • Image reconstruction
  • Parallel computing
  • Nonlinear acoustic tomographic imaging
  • Performance improvement in computational methods
  • Acceleration of inversion algorithms

Categories

Download Article