When carrying out fluorescence imaging in thicker samples, we often hear about how “scattering” affects image quality by limiting the depth of imaging within the sample. Scattering occurs when light traveling through a sample encounters a change in refractive index, thus leading to a change in the original trajectory. As cells and tissues are complex mixtures of objectives and particles with different refractive indices, this means that scattering occurs any time we image biological samples. The consequence is reduced excitation and detection efficiency because light waves spread out rather than continuing along in the direction of travel. It follows then that the thicker the sample, the more pronounced the scattering, meaning that we need to apply specialized techniques such as tissue clearing and/or multiphoton microscopy to achieve greater depths of imaging.
When working with living samples, the reality is absorption is as important as scattering. Absorption is important for effective fluorescence excitation but reduces signal when molecules other than the intended targets absorb light. For example, both hemoglobin and water absorb visible light. The plot of their absorption cross-sections as a function of wavelength is shown in the figure (A. hemoglobin, B water). It can be seen that absorption is minimized in the wavelength range between 650 and 1350 nm. This wavelength range is denoted the “near-infrared window” for imaging biological tissue because both absorption and scattering are minimized, and allows for the greatest depth of imaging within tissue.
As tissue clearing is not compatible with live samples, multiphoton microscopy is the go-to technique for thicker living preparations. Multiphoton effects are very dependent on the tightness of focus because we need many photons compressed into a small volume to make sure they are likely to encounter each other and interact. While IR lasers scatter less than visible lasers, the limiting factor boils down to how much absorption or scattering the laser excitation experiences on its way to the focal plane, again highlighting the importance of the near-infrared window for imaging live samples.
Another advantage of multiphoton lasers is that they are frequently tunable over a broad spectral range. We are often restricted to particular wavelengths because of the fluorophores we are using, but there is often some wiggle room in the choice of wavelength. Some tissues also scatter or absorb in different wavelength bands, and it may be worth seeing if there is a wavelength band where your tissue absorption and scattering are minimized, allowing for imaging deeper into a living sample.
Along with optimizing the excitation wavelength, it is important to test the image quality using different objectives when trying to image deeper into tissue. As light travels deep into a sample, the chance that it will scatter depends on how much material it passes through, as well as the angle at which it travels through any boundaries. We normally want a high numerical aperture (NA) for tight focus for multiphoton imaging. Counter-intuitively, however, a lower NA, looser focus can actually perform better in the presence of scatter. This is because the high angle rays coming out of the edge of a high NA objective are more likely to experience scattering effects. They travel through more material than the central rays and when they encounter a boundary they do so at a steeper angle, making them more prone to being deflected away from the focus.
On a closing note, a recurring theme in microscopy is that often what’s negative in one application is beneficial in another. Although absorption and scattering can create issues in fluorescence imaging of biological tissue, they are also important ways of generating contrast in optical microscopy. For example, histology is based on the absorption of light by dyes, while darkfield microscopy creates contrast through light scattering. So next time you encounter a limitation, recognize that it may be the inspiration for novel and powerful ways to visualize your sample.