Characterizing Deep Optical-Sectioning Microscopy Performance With Scattering Phantoms And Numerical Simulations

Microscopes are being developed for use in living animals, and even humans, to image microanatomical changes and molecular markers that are associated with disease. Phantoms that can be used to evaluate the performance characteristics of these systems have not been well described or standardized. We have been developing the tools to evaluate a dual-axis confocal (DAC) microscope design to optimize the features required for in vivo diagnosic imaging, and these may have features that are useful for evaluation of other such devices. We have performed diffraction-theory modeling, Monte-Carlo scattering simulations, reflectance experiments in tissue phantoms, and tissue-imaging validations. First, we determined how scattering from tissue deteriorates the diffraction-limited transverse and vertical responses in reflectance DAC imaging. Specifically, the vertical and transverse responses of the DAC to a plane reflector and a knife edge, respectively, were measured at various depths in an Intralipid scattering phantom. Comparisons were made with both diffraction-theory and Monte-Carlo scattering simulations. Secondly, as a practical demonstration of deep-tissue fluorescence microscopy, three-dimensional fluorescence images were obtained in thick human biopsy samples. These results demonstrate that the efficient rejection of scattered light in a DAC microscope enables deep optical sectioning in tissue. Finally, we will discuss our needs and plans for similar tissue-phantom experiments to validate the performance of multimodal optical-and ultrasound-imaging platforms under development. As devices are developed for the imaging of epithelial surfaces and substructures, standardized phantoms that represent the multilayered anatomical features of these tissues will need to be developed.