Look closely at a black and white photograph taken long ago at a family gathering and you will see shades of gray in everything from furniture to clothing. Is your mother’s outfit gray? Might it be blue? Or is it actually a yellow suit that happens to appear dark in the photo? In the same way that color photographs provide more information about a scene than black and white photographs, multispectral imaging can provide more information about the “color” properties of a tissue sample than a simple RGB color camera.
For many decades, pathologists have used immunohistochemical staining for protein distributions to examine tumor sections on a microscope, often one equipped with a color camera. These color cameras detect only three visible wavelengths that mimic the human eye: red, green, and blue. The human eye and color cameras work very well for looking at the single stains (typically a brown DAB stain for the protein distribution and a blue hematoxylin counterstain to visualize the tissue architecture). However, in order to look at the distributions of multiple proteins, pathologists currently must look at one microscope slide for each protein, which adds time and can only give a reading on the overall protein expression in the tumor, and not on the protein expressions in individual cells.
“And that means more ‘seat time’ for the pathologist,” explains James R. Mansfield, a leading spectral imaging scientist. “The pathologist may now have to examine as many as four or five different tissue slices under a microscope to detect and quantify multiple clinically important proteins in a single tumor sample.”
The introduction of multispectral imaging technology and new multispectral imaging systems has erased this problem by enabling researchers to spectrally resolve up to five or six chromogens (colors) in a single tissue section, even if those chromogens are spatially overlapping and co-localized.
“Multicolor labeling is common in many clinical methods such as flow cytometry, which analyzes homogenized samples,” says Mr. Mansfield, “but multicolor methods that maintain the morphology and pattern of distribution of the protein are changing the way researchers can view their samples.”
Once the sophisticated imaging system is attached to a standard microscope, researchers can stain up to four proteins using different colors and look at tissue samples with 10 to 30 different wavelengths, allowing for the accumulation of more information than is currently available. This helps researchers to better understand the complicated signaling pathways in cancer cells, and to develop more targeted therapies, which might allow physicians to better personalize treatment for individual patients.
“By combining multispectral imaging technology with sophisticated automated morphologic segmentation and image analysis methodologies,” says Mr. Mansfield, “we now have a new toolbox for the exploration of cellular phenotypes and events in solid tissues.”
Where Are They Now
Once attached to a standard microscope, this imaging system enables researchers to spectrally resolve up to five or six chromogens (colors) that are used to stain specific elements in a single tissue section and accumulate more information compared to previously existing techniques. The ability to stain for multiple cell markers substantially speeds information acquisition by providing results from a single assay that would normally require multiple tests. In addition, simultaneous staining of multiple markers allows researchers to better understand interrelationships of complicated signaling pathways in cancer cells (or other abnormal tissues), allowing the development of more effective therapies.
Novel advances in 2015 created a multispectral microscope with the ability to process 17 billion pixels representing 16 different color channels, in a single image. Using multiple microlenses to capture the image, the new system ameliorates pharmaceutical research, especially cancer research, to be performed more efficiently and faster than ever. The next goal is to create a device with the ability to capture high quality time-lapse movies of live cells, allowing researchers to preform experiments that currently aren’t possible with small scale time-lapse analysis.
In 2017, the imaging market experienced a shift toward hyper-spectral imaging systems, as hyper-spectral imaging provides greater sensitivity and specificity. Hyper-spectral imaging measures contiguous spectral bands, as opposed to multi-spectral imaging which measures spaced spectral bands.