Imaging Software Compares with Confocal Microscopy
Daniel C. McCarthy
Access to confocal microscopes can often be limited and, if there is a usage fee, costly. This and other trade-offs to confocal microscopy's high-resolution, three-dimensional images has sent some researchers hunting for supplemental analysis methods. One approach is the application of deconvolution software to digital images that were captured under standard light microscopes. Barry Condron, a neuroscience researcher at the University of Virginia Medical School, is using HazeBuster deconvolution software from VayTek Inc. in his study of neural development in embryonic grasshoppers.
Axon development in grasshopper embryos presents a viable model for growth of more sophisticated neural systems in humans. Deconvolution software applies imaging algorithms to light microscope images to mimic the function performed by the apertures in laser-scanned confocal microscopes.
Like most biomedical research, Condron's work only sounds esoteric. His lab is studying how neurons develop in the brain, specifically those neurons that respond to drugs like cocaine and Prozac or that are affected by diseases like autism. Human brains develop perhaps 25,000 of these cells, whereas grasshoppers grow only one. Examining how quickly the neuron develops and whether or not it reaches the correct synaptic connections can help explain the influence of certain drugs or disease.
"Imaging is everything for us," said Condron. "We need to watch the development of these cells and follow molecules expressed in them -- often in the live animal."
"I wanted my own [confocal] microscope," he added, "but deconvolution software made our regular light microscope close, in terms of sharpness, to a confocal instrument." VayTek's software mathematical algorithms perform the same function as the apertures in laser-scanned confocal microscopes. Namely, they remove the blur or haze contributed by out-of-focus image planes.
Instead of taking one high-resolution image of a single grasshopper axon and examining it for fine detail, Condron captures images from several insects and tabulates all the data to build a statistic. "Confocaling is very slow. The scans are slow," he said. "Using software to enhance
light microscope images, we can
just whiz through samples."
Condron pointed out that deconvolution software, including VayTek's, has trade-offs that make it a supplement to, more than a substitute for, confocal microscopes. "You're not always sure: Is this image biologically real or a product of mathematical algorithms? You need to have a confocal or some other high-resolution backup to periodically check what you are seeing."
MORE FROM PHOTONICS MEDIA