Editorial Article
Monday February 02, 2009
by Catherine Shaffer
If there's one universal need shared among all life sciences laboratories, it's the need for good data analysis tools. This is especially true in the realm of image analysis, where a single high content screen can generate gigabytes of data very quickly. A great many applications and instruments in the life sciences require image analysis, and rarely will one application meet every need. There is a growing trend toward enabling users to customize the software themselves, for their own unique needs.
Open Standards
One of the most exciting developments in image analysis software
is the advent of the open microscopy environment (OME), an open source academic collaboration to produce tools for biological light microscopy (more information). Commercial software developers and instrument vendors are beginning to pay attention to the do-it-yourself spirit of the OME by bringing their proprietary software packages into compliance with open microscopy standards. Andor
(Belfast, UK), for example, has rewritten the graphics code for the upcoming release (version 2.0) of its popular iQ image analysis software to conform with Open GL, a software standard that streamlines integration of the graphics with hardware. They have also incorporated compatibility with Python, a scripting language that is widely used in the scientific community. Python has been around since the early 90s, and is especially useful for handling large arrays of data. Using Python, scientists can write their own algorithms, or use ones that are already out in the community.
Regarding the shift toward open standards, says Mark Browne, PhD: “We believe customers should have access to data that belongs to them. It doesn't belong to the instrument manufacturer.” In the past, many scientists have been frustrated by losing access to their data, or having to go through painful contortions to convert proprietary file formats when they change operating systems—much the same thing many of us experience on a smaller scale with our personal computers. “I believe that for too long the industry relied on a closed source model ... we want to be part of this [open source] movement,” says Browne.
The new, streamlined version of iQ will be able to handle complex tasks, such as simultaneous multicamera acquisitions. Using a dual port camera, you can split an image either by wavelength or polarization, and observe both channels simultaneously at full resolution, achieving higher frame rates within the same specimen. For example, such a system is capable of simultaneous excitation and detection in a fluorescence microscopy experiment.
Have it Your Way
Often one of the unfortunate choices that must be made in selecting a software system is whether you want a system that is easy to use right out of the box, or a system that allows a great deal of flexibility and customization by an advanced user. If you have a straightforward application and little time to scale the learning curve, the out-of-the-box option can seem very attractive. But with time and experience, it's not unexpected that a basic-level user will eventually want to try some unique protocols that do not come with the software package.
GE Healthcare
is targeting this problem with its latest release of IN Cell software, which primarily focuses on high content cellular analysis. The IN Cell software suite enables automated image acquisition, analysis and data management capability for the drug discovery and development environment. The platform includes two software applications that work together to provide full functionality needed to analyze and manage high content images. IN Cell investigator is a comprehensive software package that allows the user to segment images of individual cells with a large selection of pre- and post-processing tools. IN Cell miner is an application that manages all of the information collected from high-content screens. Like Andor's iQ, IN Cell miner is based on an open architecture philosophy and is compatible with open microscopy standards. Abhay Kini, PhD, from GE Healthcare, says, “We don't want customers tied in to a vendor... the customer's data belongs to them. They are free to access their data at any time and in any form they choose to, and don't have to come back to GE Healthcare to request access."
IN Cell Miner was released in the fourth quarter of 2008, and in the second quarter of 2009, GE Healthcare will be releasing version 1.5 of the IN Cell Investigator image analysis software with a number of new capabilities. Extended focus, an image transformation protocol, enables the creation of a 2-D extended focus image from a z-series stack of images. Image stitching, another new protocol within the IN Cell investigator suite, will allow customers to create high quality panoramic images from a set of overlapping images, providing a view of a larger population of model organisms.
Compatibility
MDS Analytical Technologies
is home to one of the most mature image analysis software products on the market. Metamorph is a very stable system with a loyal following. Its main strength is its extensive hardware compatibility. Instrument drivers are difficult to get right, and being able to offer customers the assurance that the software will work with their particular system is an attractive feature. Metamorph is compatible with most third party hardware systems, and will provide virtually any sort of analysis or image processing required for biological sciences. Metamorph version 7.6 offers two new application modules: 3D motion analysis and micronuclei. The 3D motion analysis module allows you to track objects in multidimensional space, such as tracking the movements of cells or subcellular components over time. Micronuclei is an automated application module for detecting cells that are expressing micronuclei.
Version 7.6 also introduces support for a couple of additional hardware vendors, including the Ludl slide loader and the Optigrid structured illumination device, which provides a confocal-like imaging capability.
MDS puts out a major new release of Metamorph every year, based on feedback from their customer base. Says Chris Kier, PhD, director of marketing for Metamorph: "We're listening to our customers and resellers. We keep an eye on what's emerging in the research imaging community. Our real strength is our ability to react quickly to changes. We focus solely on software development, and we're able to write drivers very quickly. We build up application modules quickly too, in four weeks' time, rather than a year. Additionally, we have a real depth of application knowledge based on our 20 years of history that lets us better understand our customers' imaging needs."
Image analysis generates more data than any other type of data analysis in the life sciences. Each individual assay may generate a file of many megabytes, and a high content screen hundreds of gigabytes. It is literally impossible for one researcher or a small group to analyze all of their data manually. It's worth thinking as hard about your image analysis suite as your experiment.