German software firm Definiens this week officially announced the release of the latest version of its Cellenger software for high-content image analysis, and company officials said the firm plans to release another version of the package at the beginning of next year.
The software released this week, Cellenger v1.4, contains a set of enhancements intended to solidify the company’s strategy of offering highly flexible image-analysis software. More importantly, its performance in the market may hint at whether the market for platform-independent image-analysis software for high-content analysis can sustain itself.
The most important new features of Cellenger 1.4 include improvements to the graphical user interface, an expanded image-analysis module library, and the ability to tweak the module library to perform specific screening or imaging experiments for which Definiens may not have created algorithms.
“This is something customers are doing with the system already,” Nick Arini, Definiens’ Cellenger product manager, told CBA News this week. “They get it in, and they are immediately building on it and customizing it to meet their own problems.”
Furthermore, Definiens has expanded the list of HCS instruments with which Cellenger can be integrated, and has created more “seamless” integration to those instruments than in the past, Arini said. Currently, Definiens offers Cellenger connectors for the Becton Dickinson Pathway HT; the Cellomics’ ArrayScan; the Evotec Opera; the GE IN Cell Analyzer 1000 and 3000; the MAIA Scientific MIAS; and both the Molecular Devices Discovery-1 and the MetaXpress software, which is part of the ImageXpress instrument.
“Those are the ones that we support because they’re the most common, and we have people actively using all of those,” Arini said. He added that “a few others [have recently] come up, so we’ll need to add to that.” He declined to provide the names of specific instrument vendors or platforms.
Offering a preview of Cellenger v2.0, Arini said that one of the major upgrades will be a software tool that will allow users to basically create their own connectors to the high-content data format of their choice. “So even if you don’t have support for a particular instrument, the tool we would use to make that support is available to the user,” he said.
Jack of All Trades
In the quickly developing market for high-content image-analysis software, Definiens is attempting to carve out a niche between two extremes: highly flexible but technically demanding software that requires user programming, and so-called “pre-canned” image-analysis packages that are rigidly defined but user friendly.
“I like to think of this as a bit of a continuum,” Arini said. “At one end of the scale you’ve got a hardcore programming language such as C++. You can make image-analysis routines in pure C++. You can virtually do anything with a pure programming language.
“At the other end of the scale you have what people would call a canned analysis,” he added. “The classic example of this would be something like a Cellomics BioApplication. It can do one job, it can really only do the one job it was designed to do, and there is no way you can change it.”
Each of these extremes presents a problem for certain users. On one hand, when users of pre-canned algorithms come across a problem the algorithm isn’t designed to solve, “they actually end up designing their assays around the limitations of their image-analysis algorithm,” Arini said.
On the other hand, tools like Media Cybernetics’ ImagePro Plus, the MathWorks’ Matlab, and “to a certain extent” Molecular Devices’ MetaMorph “are very flexible and generic, but require a lot of technical chops to use them,” Arini said. “The average research scientist is just terrified of going into something like that. There’s a really big learning curve.”
The latest Cellenger release, according to Arini, is an attempt to move along this continuum toward the “user-friendly” region by enhancing the software’s user interface while retaining the programming flexibility that Definiens has always considered to be one of its strongest selling points.
Definiens released the software at Cambridge Healthtech Institute’s High-Content Analysis Europe conference in Vienna, Austria, this week, but the product had actually been on the market since September, Arini said.
And even though Definiens is focusing on selling v1.4 to existing and new customers, the company is already making plans to release v2.0 at CHI’s US-based HCA meeting, to be held in San Francisco in January 2007.
“To be honest, our external marketing is a little behind our product development group,” Arini said. “The development team is moving quite quickly, and that’s a benefit of the way that we make image-analysis solutions. We’re able to quickly turn around new technology, because we don’t have to go down to the core software development team.”
Arini further explained that Cellenger is based on a generic technology it calls Enterprise Image Intelligence. This core technology also serves as the basis for Definiens’ other software packages for applications such as satellite and aerial image classification, and histopathology and non-invasive medical imaging.
“Obviously you need to be able to take those generic tools and make them useful in a specific domain, and add specific functionality,” Arini said. In Cellenger the company takes “a very generic platform and build[s] on it in terms of specificity to cell-based assays.”
“To be honest, our external marketing is a little behind our product development group.”
Several industry insiders have in the past predicted that platform-independent image-analysis software providers would likely struggle to capture significant HCS market share, primarily because they would not be able to offer anything new for customers beyond the software that comes pre-packaged with HCS instruments.
However, it is becoming clearer that not all instrumentation vendors excel in the image-analysis arena, and that even those who do may offer software that is not necessarily conducive to all the possible applications desired by users or cross-platform compatible. Therefore, the industry will be watching closely to see if Definiens’ agnostic approach has legs.
It is unclear how successful Cellenger has been thus far, since Definiens doesn’t disclose its revenues.
According to Rene Hermes, vice president of marketing for the company’s life sciences and earth divisions, approximately half of Definiens’ software for life sciences applications is sold in North America, and half in Europe. Hermes also said that the life sciences market accounts for just over half of Definiens’ revenues, with satellite and environmental image analysis accounting for the bulk of its remaining revenues.
Definiens also plays in more than one life sciences market, which may help stabilize its revenues if the market for independent high-content image analysis software fizzles. Within the life sciences space, the company currently sells most of its software for cellular imaging applications, but also to a lesser extent for histopathological analysis of clinical tissue samples.
Furthermore, the company intends to expand its life sciences presence by moving into non-invasive radiological medical imaging.
“We play in all areas, but the main focus now is cell as well as tissue,” Hermes said. “A new area where we will start developing products now is non-invasive imaging. It’s a large market, and a tremendous fit for our technology, but it just needs some additional research to have a perfect match there.”