SVision, an image-analysis software company based in Bellevue, Wash., recently won a $750,000, Phase II Small Business Innovation Research grant from the National Institutes of Health to develop high-content, high-throughput, live-cell analysis capabilities for its SVCell software.
The grant was awarded under the “High-Throughput Tools for Brain and Behavior” program under the NIH’s National Institute for Mental Health and the National Institute of Neurological Disorders and Stroke. It marks the company’s fourth NIH SBIR grant for live-cell analysis, and is its first to reach Phase II.
Sam Alworth, business and product manager for SVision, said the company is working with eight labs on the project, although he declined to specify any of the collaborating institutions.
The company recently completed a feasibility study for Phase I of the grant in collaboration with Jane Sullivan’s lab at the University of Washington.
SVCell, which the company launched last fall, initially provided image segmentation, measurement, and classification capabilities (see CBA News, 10-13-06).
The company has been using its SBIR funding to develop new capabilities for live-cell assays including time-lapse tracking and kinetics analysis. The latest grant brings the total SBIR funding for the project to more than a million dollars.
SVCell provides interfaces for “recipe” creation, Alworth said. Scientists can use these interfaces to “teach” the computer to automatically detect, segment, measure, classify, and analyze cellular and subcellular phenotypes in their microscopy imaging experiments, he said.
“The focus of the current project is on commercialization of this subcellular time-lapse module and having it work with our teaching interfaces in SVCell,” Alworth said. “Right now we have a module that works for an assay, but we have to connect it to the teaching module in SVCell, so that it will work with lots of similar assays.”
Alworth told Cell-Based Assay News that the NIH “hooked the company up” with its network of high-throughput screening centers, and SVision is working with at least one screening center to test the throughput capabilities of SVCell.
He declined to name the screening center, however, noting that the agreement is still “very informal.”
The company plans to hire more staff, Alworth said, though not so much to work on the grant as to work on application development for SVCell. The company’s core team of scientists and software engineers are working on the SBIR-funded project.
Alworth explained that the Phase II SBIR grant focuses on the analyses of puncta, or small structures, such as quantum dots. It could be that the assay needs to identify the small structure, or that scientists are looking to analyze a signal, such as an ion fluctuation, through the small structure.
The company released SVCell 1.0, the first version of SVCell with live-cell imaging capability, in May, but this version did not include tracking capability.
That feature will be included in the 1.1 version, which is due out in November, Alworth said.
Alworth said that SVision’s business focus is to find distribution partners, co-marketing partners, or original equipment manufacturer partners. “We hope to be able to announce those partnerships later this year,” he said.
In terms of Phase II project milestones, Alworth said that there will initially be a lot of technology development around the teaching interfaces, so people can easily create the new analyses that they need.
He said the second stage will probably be more focused on doing experiments with the collaborators for the necessary validation of the tool, the software, and the engineering.
“The focus of the current project is on commercialization of the subcellular time-lapse module and having it work with our teaching interfaces in SVCell.”
“As the components of the system mature, they will be brought into the software,” said Alworth.
The NIH has been integral in helping SVision develop the next generation of SVCell, which extends the machine learning concepts of identifying objects and tracking them on the front end, and analyzing biological phenomenon over time at the back end, said Alworth.
He added that analyzing time-lapse data is very difficult, but SVision’s pattern recognition technology can help people see types of movement rather than lots of kinetic data. The NIH funding will help SVision create a next-generation tool for what he called “quantitative microscopy movies.”
SVision appears to be in sync with the market. The industry has been gradually shifting from predefined inflexible image processing and analysis algorithms developed by image processing experts to image processing software environments that empower the cell biologist, Tom Moran, director of the Imaging Collection at Accelrys, told CBA News in an e-mail.
“We are encouraged to see the award made to SVCell,” Moran said. SVision’s “teach by example” interface for image segmentation allows users to incorporate their expertise in biology in a user-friendly environment while creating custom algorithms, he said.
Moran said that Accelrys has been very successful with the image processing environment for its Pipeline Pilot software, the latest version of which was launched last December. The company has incorporated both learning components as well as classical image processing components into one environment to empower image processing experts and non-experts alike.
Moran said that Accelrys is currently completing data integration projects with several of the “major image acquisition vendors,” though he declined to disclose the names of these partners.