Related Experiment Video
Updated: Jul 16, 2026
Cryo-Electron Microscopy Screening Automation Across Multiple Grids Using Smart Leginon
Published on: December 1, 2023
Architecture and performance of a Grid-enabled lookup-based biomedical optimization application: light scattering
Renato J Figueiredo1, Vadim Backman, Yang Liu
1Advanced Computing and Information Systems (ACIS) Laboratory, Department of Electrical and Computer Engineering, University of Florida, Gainesville, FL 32611, USA. renato@acis.ufl.edu
This study introduces GridLSS, a high-performance computing approach for light scattering spectroscopy (LSS) enabling noninvasive precancerous tissue diagnosis. The system efficiently analyzes spectral data using precomputed databases and optimized data transfers for improved in vivo diagnostics.
Area of Science:
- Biomedical Optics
- Computational Science
- High-Performance Computing
Background:
- Light scattering spectroscopy (LSS) offers noninvasive in vivo diagnosis of human epithelium.
- Distinguishing precancerous changes from healthy tissue requires quantitative analysis of spectral data.
- Traditional methods often involve invasive biopsies, necessitating advanced diagnostic tools.
Purpose of the Study:
- To present a Grid-enabled implementation of LSS (GridLSS) for enhanced precancerous tissue analysis.
- To describe the architecture and integration of GridLSS with a web-based Grid computing portal.
- To optimize the determination of light scattering spectra for improved diagnostic accuracy.
Main Methods:
- Developed a novel parallel application for LSS analysis within a distributed computing environment.
- Utilized rigorous Mie theory for analytical computation of spectra.
- Implemented a novel approach using precomputed spectra stored in lookup databases for efficient querying.
- Integrated the application with a web-based Grid computing portal for accessibility.
Main Results:
- Achieved high performance for LSS analysis in a wide-area distributed computing environment.
- Evaluated and analyzed application performance in typical Grid deployment cluster environments.
- Demonstrated significant performance benefits from on-demand Grid data transfers and user-level caches.
Conclusions:
- GridLSS provides a high-performance, parallelized solution for light scattering spectroscopy analysis.
- The use of precomputed spectra and optimized data transfer mechanisms enhances diagnostic efficiency.
- This approach facilitates advanced in vivo, noninvasive diagnosis of precancerous changes in human epithelium.
