Advertisement

A High-performance Computation Method for Simulation of Cardiac Excitation Propagation using a Supercomputer

  • Tohru Suzuki
  • Takashi Ashihara
  • Masashi Inagaki
  • Tsunetoyo Namba
  • Takanori Ikeda
  • Kazuo Nakazawa

Summary

Recently, attention has been focused on the simulation of the propagation of cardiac excitation as a useful method to understand the mechanism of arrhythmia. However, precise simulation based on the ion channel characteristics of cardiomyocytes requires an extremely high level of computational power due to the requirement for the numerical analysis of a differential equation system with many variables, including an exponential function. The high-performance computation method we employed, using a supercomputer, made it possible for us to simulate cardiac excitation propagation on the basis of an ion channel model within a practical time period. In this method, we used vectorization in the main high-speed architecture of the supercomputer. We also constructed and employed a piecewise linearization table for the approximation of functions and transformed the ventricular morphology into a one-dimensional array. The use of vectorization increased computation speed by a factor of about 18 times. The combined use of parallel processing further increased the computation speed 7 fold. The use of approximations employing the piecewise linearization table and the one-dimensional re-arrangement both realized independent increases in computation speed by factors of 5 times. Although it would normally require one and a half months to finish these computations with the latest personal computer, due to the vast numerical data involved, the use of the supercomputer and this method allowed us to complete the computations in only one day. Thus, innovative use of the supercomputer provides the means to perform precise and large-scale simulation of cardiac excitation propagation and we fully expect that this will bring about great advances in cardiac electrophysiology.

Keywords

Parallel Processing Piecewise Linearization Spiral Wave Differential Equation System Excitation Propagation 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Ashihara T, Namba T, Ito M, Kinoshita M, Nakazawa K (1999) The dynamics of vortex-like reentry wave filaments in three-dimensional computer models. J Electrocardiol 32 (suppl): 129 - 138CrossRefGoogle Scholar
  2. Ashihara T, Suzuki T, Inada H, Namba T, Nakazawa K (1998) Dynamics of spiral waves and the simulated ECGs in numerical model of the heart (in Japanese). The 18th Joint Conference on Medical Informatics, pp 162 - 163Google Scholar
  3. Luo CH, Rudy Y (1991) A model of the ventricular cardiac action potential: depolarization, repolarization, and their interaction. Circ Res 68: 1501 - 1526CrossRefGoogle Scholar
  4. NEC Corporation (1995) In: Outline manual of NEC supercomputer SX-4 Series, GUZ 411 (in Japanese). NEC, Tokyo, pp 8–29,209–215Google Scholar
  5. NEC Corporation (1995) In: SUPER-UX Guidebook of C programming, G1AF02–7 (in Japanese). NEC, Tokyo, pp 31–55,71–162Google Scholar
  6. Nagashima S, Tanaka Y (1992) In: The Institute of Electronics, Information and Communication Engineers (ed) Supercomputer (in Japanese). Ohmsha, Tokyo, pp 1 - 97Google Scholar
  7. Nakazawa K, Namba T, Suzuki R (1999) Dynamics of spiral waves in three dimensional FHN model media (in Japanese). Jpn J Med Electro Biol Eng 37: 63–77 (*1)Google Scholar
  8. Nakazawa K, Namba T, Suzuki T, Suzuki R (1997) A simulation study of fibrillation on FHN-model configuration of one million cells (in Japanese). Tech Repo IEICE (MBE) 97: 57 - 64Google Scholar
  9. Nakazawa K, Suzuki R (1997) Breakups and interactions of spiral waves on two dimensional FHN model, related to the mechanism of transition from tachycardia to fibrillation (in Japanese). Jpn J Med Electro Biol Eng 35: 354 - 364Google Scholar
  10. Nakazawa K, Suzuki T, Inada H, Kamo T, Doi S, Kumagai S, Namba T, Ashihara T, Suzuki R (1998) Dynamics analysis of spiral waves in the ventricular form media (in Japanese). Tech Repo IEICE (MBE) 98: 9 - 16Google Scholar
  11. Nakazawa K, Suzuki T, Inada H, Namba T, Ashihara T, Suzuki R (1999) Numerical analysis of spiral wave reentry by supercomputer (in Japanese). Jpn J Med Electro Biol Eng 37 (suppl): 467Google Scholar
  12. Nakazawa K, Suzuki T, Namba T, Ashihara T (1998) A fundamental study of spiral wave reentry by mathematical models (in Japanese). The 18th Joint Conference on Medical Informatics, pp 160 - 161Google Scholar
  13. Nakazawa K, Suzuki T, Namba T, Ashihara T (1998) A fundamental study of spiral wave reentry by mathematical models (in Japanese). The 18th Joint Conference on Medical Informatics, pp 160 - 161Google Scholar
  14. Nakazawa K, Suzuki T, Namba T, Ashihara T, Doi S, Nagata S, Inada H, Suzuki R (1999) Visualization of arrhythmic excitation propagation in the heart by spiral wave theory (in Japanese). Nikkei Science 10: 112 - 113Google Scholar

Copyright information

© Springer Japan 2000

Authors and Affiliations

  • Tohru Suzuki
    • 1
  • Takashi Ashihara
    • 2
  • Masashi Inagaki
    • 3
  • Tsunetoyo Namba
    • 4
  • Takanori Ikeda
    • 5
  • Kazuo Nakazawa
    • 1
  1. 1.Department of EpidemiologyNational Cardiovascular Center Research InstituteFujishiro-dai, SuitaJapan
  2. 2.First Department of Internal MedicineShiga University of Medical ScienceSeta Tsukinowa-cho, OtsuJapan
  3. 3.Department of Cardiovascular DynamicsNational Cardiovascular Center Research InstituteFujishiro-dai, SuitaJapan
  4. 4.Department of Cardiovascular MedicineOkayama University Medical SchoolShikatacho, OkayamaJapan
  5. 5.Third Department of Internal Medicine, Ohashi HospitalToho UniversityOhashi, Meguro-ku, TokyoJapan

Personalised recommendations