Adaptive optics applied to coherent anti-Stokes Raman scattering microscopy

We report on the use of adaptive optics in coherent anti-Stokes Raman scattering microscopy (CARS) to improve the image brightness and quality at increased optical penetration depths in biological material. The principle of the technique is to shape the incoming wavefront in such a way that it counteracts the aberrations introduced by imperfect optics and the varying refractive index of the sample. In recent years adaptive optics have been implemented in multiphoton and confocal microscopy. CARS microscopy is proving to be a powerful tool for non-invasive and label-free biomedical imaging with vibrational contrast. As the contrast mechanism is based on a 3rd order non-linear optical process, it is highly susceptible to aberrations, thus CARS signals are commonly lost beyond the depth of ~100 μm in tissue. We demonstrate the combination of adaptive optics and CARS microscopy for deep-tissue imaging using a deformable membrane mirror. A random search optimization algorithm using the CARS intensity as the figure of merit determined the correct mirror-shape in order to correct for the aberrations. We highlight two different methods of implementation, using a look up table technique and by performing the optimizing in situ. We demonstrate a significant increase in brightness and image quality in an agarose/polystyrene-bead sample and white chicken muscle, pushing the penetration depth beyond 200 μm.