Advertisement

Heterogeneous Platform Programming for High Performance Medical Imaging Processing

  • Renan Sales Barros
  • Sytse van Geldermalsen
  • Anna M. M. Boers
  • Adam S. Z. Belloum
  • Henk A. Marquering
  • Silvia D. Olabarriaga
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8374)

Abstract

Medical imaging processing algorithms can be computationally very demanding. Currently, computers with multiple computing devices, such as multi-core CPUs, GPUs, and FPGAs, have emerged as powerful processing environments. These so called heterogeneous platforms have potential to significantly accelerate medical imaging applications. In this study, we evaluate the potential of heterogeneous platforms to improve the processing speed of medical imaging applications by using a new framework named FlowCL. This framework facilitates the development of parallel applications for heterogeneous platforms. We compared an implementation of region growing based method to automated cerebral infarct volume measurement with a new implementation targeted for heterogeneous platforms. The results of this new implementation agree well with the original implementation and they are obtained with significant speed-up comparing to the sequential implementation.

Keywords

dataflow framework heterogeneous computing heterogeneous platforms medical imaging processing OpenCL parallel programming 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Augonnet, C., Thibault, S., Namyst, R., Wacrenier, P.A.: Starpu: a unified platform for task scheduling on heterogeneous multicore architectures. Concurrency and Computation: Practice and Experience 23(2), 187–198 (2011)CrossRefGoogle Scholar
  2. 2.
    Barak, A., Ben-Nun, T., Levy, E., Shiloh, A.: A package for opencl based heterogeneous computing on clusters with many gpu devices. In: 2010 IEEE International Conference on Cluster Computing Workshops and Posters (CLUSTER WORKSHOPS), Heraklion, pp. 1–7 (2010)Google Scholar
  3. 3.
    Boers, A., Marquering, H., Jochem, J., Besselink, N., Berkhemer, O., van der Lugt, A., Beenen, L., Majoi, C.: Automated cerebral infarct volume measurement in follow-up noncontrast ct scans of patients with acute ischemic stroke. American Journal of Neuroradiology (2013)Google Scholar
  4. 4.
    van Geldermalsen, S.: Work In Progress Thesis - FlowCL. Master’s thesis, University of Amsterdam (2013)Google Scholar
  5. 5.
    Khronos OpenCL Working Group: The opencl specification (2012)Google Scholar
  6. 6.
    Kothari, R.U., Brott, T., Broderick, J.P., Barsan, W.G., Sauerbeck, L.R., Zuccarello, M., Khoury, J.: The abcs of measuring intracerebral hemorrhage volumes. Stroke 27(8), 1304–1305 (1996)CrossRefGoogle Scholar
  7. 7.
    MATLAB: version 8.0.0.783 (R2012b). The MathWorks, Inc., Natick, Massachusetts (2012)Google Scholar
  8. 8.
    Pham, D.L., Xu, C., Prince, J.L.: Current methods in medical image segmentation. Annual Review of Biomedical Engineering 2(1), 315–337 (2000)CrossRefGoogle Scholar
  9. 9.
    Saver, J.L., Johnston, K.C., Homer, D., Wityk, R., Koroshetz, W., Truskowski, L.L., Haley, E.C., et al.: Infarct volume as a surrogate or auxiliary outcome measure in ischemic stroke clinical trials. Stroke 30(2), 293–298 (1999)CrossRefGoogle Scholar
  10. 10.
    Spafford, K., Meredith, J., Vetter, J.: Maestro: Data orchestration and tuning for opencl devices (2010)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • Renan Sales Barros
    • 1
  • Sytse van Geldermalsen
    • 2
  • Anna M. M. Boers
    • 1
    • 3
    • 4
  • Adam S. Z. Belloum
    • 2
  • Henk A. Marquering
    • 1
    • 3
  • Silvia D. Olabarriaga
    • 1
  1. 1.Biomedical Engineering & Physics, Academic Medical CenterUniversity of AmsterdamAmsterdamThe Netherlands
  2. 2.Department of Computational ScienceUniversity of AmsterdamAmsterdamThe Netherlands
  3. 3.Department of Radiology, Academic Medical CenterUniversity of AmsterdamAmsterdamThe Netherlands
  4. 4.University of TwenteEnschedeThe Netherlands

Personalised recommendations