To create high-resolution 3D images of tissues such as the brain, researchers often use two-photon microscopy, which involves directing a high-intensity laser at the sample to induce fluorescence excitation. However, scanning deep in the brain can be difficult because light scatters tissue as it sinks, blurring images.
Two-photon imaging is also time consuming, as it usually requires scanning individual pixels one by one. A team of researchers from MIT and Harvard University have now developed a modified version of two-photon imaging that can image deeper into tissue and perform imaging much faster than before.
This type of imaging could allow scientists to get high-resolution images of structures such as blood vessels and individual neurons in the brain faster, the researchers say.
“By modifying the laser beam entering the tissue, we have shown that we can go further and that we can do finer imaging than previous techniques”, explains Murat Yildirim, researcher at MIT and one of the authors of the new study.
MIT graduate student Cheng Zheng and former postdoctoral fellow Jong Kang Park are the lead authors of the paper, which appears today in Scientists progress. Dushan N. Wadduwage, a former MIT postdoctoral fellow who is now a John Harvard Distinguished Science Fellow in Imaging at the Center for Advanced Imaging at Harvard University, is the lead author of the article. Other authors include Josiah Boivin, a post-doctoral fellow at MIT; Yi Xue, a former MIT graduate student; Mriganka Sur, Newton’s professor of neuroscience at MIT; and Peter So, professor of mechanical and biological engineering at MIT.
Two-photon microscopy works by projecting an intense beam of near-infrared light at a single point in a sample, inducing the simultaneous absorption of two photons at the focal point, where the intensity is the highest. This long-wavelength, low-energy light can penetrate deeper into tissue without damaging it, allowing imaging below the surface.
However, the two-photon excitation generates fluorescence images and the fluorescent signal is in the visible spectral region. When imaging deeper into tissue samples, fluorescent light scatters more and the image becomes blurry. Imaging many layers of tissue is also time consuming. Using wide-field imaging, in which an entire plane of tissue is illuminated at once, can speed up the process, but the resolution of this approach is not as high as that of point-to-point scanning.
The MIT team wanted to develop a method that would allow them to image a large tissue sample at one time, while still maintaining the high resolution of point-to-point scanning. To achieve this, they found a way to manipulate the light they shine on the sample. They use a form of wide-field microscopy, projecting a plane of light onto tissue, but alter the amplitude of the light so that they can turn each pixel on or off at different times. Some pixels are lit while neighboring pixels remain dark, and this predefined pattern can be detected in the light scattered by the fabric.
“We can turn every pixel on or off with this kind of modulation,” Zheng explains. “If we turn off some spots it creates space around each pixel, so now we can know what’s going on in each of the individual spots.”
Once the researchers get the raw images, they reconstruct each pixel using a computer algorithm they created.
“We control the shape of the light and we get the response from the tissue. From these answers, we try to determine the type of tissue dispersion. When we do the reconstructions from our raw images, we can get a lot of information that you can’t see in the raw images, ”Yildirim explains.
Using this technique, the researchers showed that they could image about 200 microns deep in slices of muscle and kidney tissue, and around 300 microns in the brains of mice. It’s about twice as deep as it was possible without this patterned excitement and computer reconstruction, Yildirim says. The technique can also generate images about 100 to 1000 times faster than conventional two-photon microscopy.
This type of imaging should allow researchers to get high-resolution images of neurons in the brain, as well as other structures such as blood vessels, more quickly. Imaging blood vessels in the brains of mice could be particularly useful in learning more about how blood flow is affected by neurodegenerative diseases such as Alzheimer’s disease, Yildirim explains.
“All studies of blood flow or the morphology of blood vessel structures are based on two- or three-photon point scanning systems, so they’re slow,” he says. “Using this technology, we can really do high-speed volumetric imaging of blood flow and blood vessel structure to understand changes in blood flow. “
The technique could also lend itself to measuring neuronal activity, by adding voltage-sensitive fluorescent dyes or fluorescent calcium probes that light up when neurons are excited. It could also be useful for analyzing other types of tissue, including tumors, where it could be used to help determine the edges of a tumor.
The research was funded by the National Institutes of Health, including the National Institute of Biomedical Imaging and Bioengineering P41 program and the NIBIB Pathway to Independence Award, the Hamamatsu Corporation, the Samsung Advanced Institute of Technology, the Singapore-MIT Alliance for Research and Technology (SMART), the Center for Advanced Imaging at Harvard University and the John Harvard Distinguished Science Fellowship Program.