Architecture and performance of a grid-enabled lookup-based biomedical optimization application: Light scattering spectroscopy

Renato J. Figueiredo*, Vadim Backman, Yang Liu, Jithendar Paladugula

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

This paper presents a case study of a Grid-enabled implementation of light scattering spectroscopy (LSS). The LSS technique allows noninvasive detection of precancerous changes in human epithelium, differentiating from traditional biopsies by allowing in vivo diagnosis of tissue samples and quantitative analyses of parameters related to cancerous changes via numerical techniques. This paper describes the architecture of GridLSS and its integration with a Web-based Grid computing portal. GridLSS solves an optimization problem of determining the light scattering spectrum that best fits experimental spectral data among a large set of spectra computed analytically using rigorous Mie theory. The novel approach taken in this paper is based on the precomputation and storage of Mie theory spectra in lookup databases that are queried during the minimization process. The paper makes three important contributions: 1) it presents a novel parallel application for LSS analysis that delivers high performance in wide-area distributed computing environment; 2) it evaluates and analyzes the performance of this application in cluster-based high-performance computing environments that are typical of Grid deployments; and 3) it shows that the performance of GridLSS benefits significantly from the use of on-demand Grid data transfers based on virtualized distributed file systems and from user-level caches for remote file system data.

Original languageEnglish (US)
Pages (from-to)170-178
Number of pages9
JournalIEEE Transactions on Information Technology in Biomedicine
Volume11
Issue number2
DOIs
StatePublished - Mar 1 2007

Keywords

  • Biomedical applications
  • Grid computing
  • Spectroscopy
  • Virtualization

ASJC Scopus subject areas

  • Biotechnology
  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Architecture and performance of a grid-enabled lookup-based biomedical optimization application: Light scattering spectroscopy'. Together they form a unique fingerprint.

Cite this