The goal of this project is to develop and commercialize next generation live cell, time-lapse microscopy image recognition software specialized for high throughput quantification of subcellular functions. The software integrates novel and robust methods of subcellular time-lapse analysis and modeling not available in the current informatics tools for the enhancement of signal detection and noise immunity to significantly improve on assay throughput, accuracy, efficiency, and reliability. These include three levels of analysis tools: 1) robust object detection enhances object detection by making a non-binary, probabilistic association of pixels to objects using confidence maps; 2) robust feature optimization enhances quantitative feature measurement by utilizing the spatial temporal information within the entire image or movie to refine the confidence maps or weight them for model fitting; and 3) outcome directed model fitting enhances the assay model parameter through iterative fitting with built-in reliability assessment and error correction using spatial-temporal image information The goal of this phase II project is to incorporate these technologies into the SVCell(tm) platform, generalize and characterize their performance in a wider range of new subcellular assays, and prepare the entire platform for product release through our commercial partners. The phase II deliverables will be a market ready software package for basic and drug discovery scientists.
Our specific aims are: 1) Optimize and validate the subcellular analysis module for broad bio-assay application; 2) Product software engineering of the subcellular analysis module in SVCell; and 3) Evaluate the product readiness of the SVCell beta through field tests and scientific collaborations. Imaging assays looking at subcellular phenotypes in both fixed and live cells are at the cutting edge of life science imaging research. They provide researchers with new tools to dissect the mechanisms of cellular function with great resolution. They lead to new insights and new discoveries that can have significant impact across all of basic research. These new assays can be rapidly scaled and translated into imaging screens for drug or biological discovery and disease diagnosis using the SVCell platform. Overall this phase II project promises to make a significant impact on human health by increasing the speed and efficiency of basic research, high throughput imaging assay development, and deployment of novel high throughput imaging assays with subtle phenotypes. Microscopy image recognition software promises to make a significant impact on human health by increasing the accuracy, speed and efficiency of basic research, high throughput imaging assay development, and deployment of novel high throughput imaging assays with subtle phenotypes. It will provide researchers with a new tool to dissect the mechanisms of cellular function with great resolution. ? ? ?

Agency
National Institute of Health (NIH)
Institute
National Institute of Mental Health (NIMH)
Type
Small Business Innovation Research Grants (SBIR) - Phase II (R44)
Project #
6R44MH075498-03
Application #
7477872
Study Section
Special Emphasis Panel (ZRG1-MDCN-L (10))
Program Officer
Grabb, Margaret C
Project Start
2005-08-01
Project End
2010-07-31
Budget Start
2008-08-01
Budget End
2010-07-31
Support Year
3
Fiscal Year
2008
Total Cost
$354,723
Indirect Cost
Name
Drvision Technologies, LLC
Department
Type
DUNS #
827582656
City
Bellevue
State
WA
Country
United States
Zip Code
98008
Alworth, Samuel V; Watanabe, Hirotada; Lee, James S J (2010) Teachable, high-content analytics for live-cell, phase contrast movies. J Biomol Screen 15:968-77