An R&D 100 Award-winning microscopy system enhances its sleuthing credentials with an integrated hyperspectral imaging system.
In 2005, a young microscopy firm called CytoViva, Auburn, Ala. set out to solve a brand-new problem. They wanted to give live-cell researchers a tool that better informed them of the way fluorescing targets interacted with other, more mysterious objects.

The solution was a dual-mode fluorescence (DMF) system that put both scanning and optical modes within the same viewing picture. Only a twist of the dial was needed to pinpoint the fluorescent targets then immediately witness the activity real-time.

The DMF module was immediately appreciated for its effective simplicity, so much so that many wondered why such a system had never before appeared.

Breast Cancer Cells
A major U.S. cancer research center is preparing a protocol for examining tissue biopsy with the HSI system. Image: CytoViva
However, the DMF presented a new problem. Not used to having high optical resolution in conjunction with fluorescence, researchers were seeing nanoscale features that had previously escaped their notice. Inevitably, interest in these smaller objects grew.

In consideration of this renewed interest for better details—and the prospect of more applications for their systems—CytoViva began building a new standard, the CytoViva Hyperspectral Imaging System (HSI).

Recently launched, the HSI now provides spectral analysis of materials and biologicals imaged with the company’s existing optical microscopy products. This capability, provided in the visible near-infrared spectrum and backed by highly developed software, has given the microscope user a powerful computational tool designed to complement high-power fluorescent imaging. Instead of one screen of information, researchers now have two.

In this respect, observers of this new spectral data now have something in common with an unlikely benefactor: the U.S. military.
From environmental chamber to spectrophotometers
In recent years, the U.S. Dept. of Defense has developed tremendously accurate systems for space-to-ground analysis. Google Earth provides a small glimpse of this new capability. The technology CytoViva is leveraging for its system is directly sourced from the same kinds of tools used to find tanks or planes on the ground from space.

“We’re pretty proud of getting a pixel to represent about 60 nm,” says Chuck Ludwig, president of CytoViva. “The satellites they have they are able to get down to 12 square inches.”

It’s a whole different scale, Ludwig continues, but it’s a similar ballgame.

The story begins with CytoViva’s 2005 introduction of an optical illumination system (the CytoViva) that features a 24 W metal halide light source to provide strong collimated light with low noise levels. This formed the basis of the DMF in 2006, which could be used in concert with most existing upright or inverted optical microscopes.

Live-cell observation was made possible by a perfusion chamber developed by Warner Instruments, Hamden, Conn., which mounts on the stage where the slide is typically mounted. This is considered a microfluidics chamber, says Ludwig, and is designed to support cellular and polymer experimentation. Essentially, three ways of observing samples were available to users:

• fluorescence-only imaging
• simultaneous fluorescence and non-fluorescence imaging; or
• high-resolution, non-fluorescence imaging.

Though the DMF system was successful after it was launched in mid-2006, its developers were peppered with questions about the identity of the features they were now seeing. The system achieved resolutions under 100 nm and detection levels under 50 nm, which, for life scientists in particular, revealed objects that many researchers hadn’t observed before.

With its Hyperspectral Imaging System (HSI), CytoViva engineers designed three major sub-systems to interact seamlessly with the existing nanoscale optical microscope and its DMF module:

• concentric imaging spectrophotometer
• customized hyperspectral image analysis software; and
• an integrated motorized stage

At the top of the HSI system, CytoViva has moved its imaging camera to a dual-port optical mount that is essentially a split light path. This works 80% on imaging mode so that users can move around the slide to find the most interesting part of the sample to study.

A small rod is pushed in to divert the light path to a concentric imaging spectrophotometer containing original, aberration-corrected convex holographic diffraction gratings. The patented module, developed by Headwall Photonics, Inc., Fitchburg, Mass., a division of Agilent Technologies, was developed to provide exceptional signal-to-noise ratio (SNR) to ensure that only spectra from the high-contrast CytoViva microscope image is captured for analysis. The system collects data in the 400 to 1,000 nm range.

Ludwig’s company has modified the hardware to fit on the microscope, altering the software to work with very high SNR ratio.

Anthrax Spores
Hyperspectral imaging of anthrax spores. Image: CytoViva
“It turns out that it’s a lot easier to take a very high quality, high signal-to-noise darkfield image than it is flying over in a plane, shooting through clouds and trying to catch something that’s reflected by the sun,” says Ludwig. “We have the optical system resolved to 100 nm and can detect down to 30-50 nm. We can detect spectral images down to 10-15 nm. [When light] backscatters off, we capture off the particle, then we subtract the background using a spectral map, which gives you a pure signal.”

But how is the data collected? A motorized stage, integrated onto the CytoViva microscope system, acts as a“push broom,” moving the sample over the hyperspectral image detector. The motion is controlled by the software, and features a range of 114 mm x 75 mm. The step size of the stage is 0.01 mm, which helps allow for a spectral resolution of 1.29 nm at Nyquist limit with a 12.5 mm slit.

To give an example, if a live epithelial cell incubated with 100 nm gold particles is imaged by the CytoViva microscope, the scan stage moves in two dimensions and various z levels.

“It goes across the x pixel by pixel, scans the array, and stores the spectral data in that pixel. So it goes to the top, like a deck of cards and creates a data cube. The data cube is a very large set of data you can return to and analyze,” says Ludwig. Then the software takes over.
Computational imaging finishes the job
The ITT Visual Information Solutions ENVI hyperspectral image analysis system was originally developed to easily and quickly interpret and analysis spectral data collected by satellite cameras. It builds on three basic ideas, built up by the military in the past 12 to 14 years, says Ludwig:

• Target identification. The military builds what they call “ground truth” (or a library sample) of stored features, such as tanks or planes. As they are scanning over a field, like a sample of tissue, the software identifies the target.

• After this development came target recognition. An automated module designed in IDL language, which is versatile and can be tweaked, pinpoints areas of interes

• The anomaly detection tool. This is probably the most advanced feature.

For example, if a satellite scans over an airfield and finds a plane or an object that should be in the library but is not, the system identifies the anomaly and marks for later examination. In microscopy, this is useful for locating strange nanoparticles. According to Ludwig, the spectrophotometer used in the CytoViva HSI is almost the same as that used in the hyperspectral imaging satellites. As part of this functionality, a zoom mode acts as a set of crosshairs, honing in on a region of interest. This mode is flexible, allowing various inclusion shapes, such as ellipses or rectangles, and criteria such as spectral intensity. Users can pick any location in the cell that conforms to a specific spectral profile, such as a drug attached to a nanoparticle, and color it to track the signatures as they change. The kinetics of drug delivery can be observed.

However, Cytoviva has honed the primary functionality of HSI, which is to easily locate regions within a sample that match up with qualified spectra that have been preloaded in a library. It has also been optimized to untangle spectra data in order to pick out specific sample components (down to a single pixel) or evaluate anomalies. Typical scans can be collected in about five minutes.
Hyperspectral tools likely to find many uses
Imaging science has made great gains in recent years; we have seen the pairing of traditional microscopy techniques with sophisticated algorithms and powerful computers, we have witnessed the driving force of nanotechnology, and we have seen—in the case of CytoViva and other systems—the combination of proven technologies for the creation of powerful new systems. Examples include the pairing of confocal Raman with transmission electron microscopy (TEM) and the adaptation of high-speed camera imaging with TEM.

CytoViva is the first to join fluorescent imaging with hyperspectral imaging in a system already known for its live-cell dual mode fluorescence approach. While this system was specifically designed to support nanoscale studies, it also supports a wide range of biomedical research initiatives in areas such as cell biology and infectious disease.

As a result, the company envisions wide appeal for HSI in the marketplace. Oncology researchers are now attempting to use nanoparticles as sensors to detect malignancy or as vehicles to bind onto cancer molecules or even enter soft tumor cells and deliver drugs, according to Ludwig, and a major U.S. cancer research center is preparing a protocol for examining tissue biopsy with the HSI system.

A more direct and immediate application has been seen in the study of anthrax.

“They are looking for a way to capture signature of the spores and the spore sleeve to be able to identify where it came from and hyperspectral imaging is a very appropriate way of doing this,” says Ludwig.

Another project with which CytoViva has been involved at the U.S. Dept. of Agriculture is in the study of C. elegans, which is a worm that has a well developed neurological system suitable for study because it responds in a relatively predictable way. They want to look at the neural sensors and be able to capture the chemical signature of this organism in certain reactions.

The USDA has ordered a system for that kind of work, and another HSI system was ordered by the U.S. Food & Drug Administration for use in nanotoxicology to monitor the effect of certain nanoparticles, such as titanium dioxide, on skin tissue.

Still other applications may emerge as nanotech and biotech continue their strong growth.