Research
MicroSeer, Analysis software for microscopy imagery
Dunrie A. Greiling, BioMedware, Inc., PI
This work was supported by an SBIR phase I grant for $177,259 from the National Institute of Biomedical Imaging and Bioengineering (R43-EB000575).
Research Abstract
This project will create new pattern recognition software to improve the analysis and interpretation of in vivo biomedical imagery. Currently, researchers can get remarkably detailed images of living cells and their constituent proteins using molecular genetic and microscopy-based approaches in conjunction with sophisticated microscopy hardware. Available image analysis techniques and software, however, lag behind the power of this new imaging equipment to visualize the microscopic world.
This phase I SBIR project will apply existing technology in spatial analysis of satellite image data to microscopy data, create new statistical techniques specific to the study of spatial association of proteins in cells, and create software that implements these statistics for use in the analysis of spatial association in timeslice in vivo biomedical imagery.
Conference
BioMedware held a conference in April 2003 for planning and review of the MicroSeer project.
Screenshot from the prototype software.
Sample image at left with two randomizations
toroidal shift in the middle and CSR at right
Prototype Software
BioMedware produced a prototype software product that was included with the Phase II grant application. The screenshot at right shows a sample image alongside two randomized versions (the middle image has one channel shifted, while the rightmost image has the green channel completely randomized).
Presentations
Greiling, D.A., W. Bement, and R.G. Rommel. 2003. Towards Objective Colocalization Statistics: Monte Carlo Analysis Using Shifts in Channel Registration. Poster presented at the American Society for Cell Biology Annual Meeting, San Francisco, December 2003.
Greiling, D.A., W. Bement, and R.G. Rommel. 2003. MicroSeer: Software for Monte Carlo evaluation of colocalization statistics using shifts in channel registration. Poster presented at the Biomedical Information Science and Technology Initiative conference in Bethesda, MD, October, 2003.