Advances in laser-scanning optical microscopy, a high-resolution imaging technique, have made it possible to image the chemical and structural details of living cellular and tissue systems at high resolution. Unfortunately, microscopic variations in tissue composition scatter the laser light, causing degradation of the image quality and limits in imaging depth. In fact, light scattering is the main limitation of using laser-scanning microscopy for tissue imaging. A promising approach to improve image quality and depth is to use adaptive optics (e.g. deformable mirrors) to shape the laser beam in a way that counteracts the effects of light scattering in the tissue. While promising, adaptive optics techniques currently use optimization schemes that do not account for the mechanisms of scattering. This leaves the method prone to artifacts, especially as scattering becomes more severe at greater depths. Thus, a mechanistic (physical) insight into the scattering would be very helpful to guide and fundamentally improve adaptive optics optimization schemes. The goal of this project is to develop an integrated modeling, computational, and experimental approach--a Virtual Microscopy Simulator (VMS)--to determine the adaptive optics signal corrections necessary to fully correct for scattering induced distortions relevant to laser scanning microscopy. The outcomes of this work improve adaptive optics technics and allow imaging at greater depths in tissue materials, which will dramatically improve the impacts of optical microscopy in the biomedical sciences. Beyond making the VMS platform available to the science community, Education and Outreach plans include participating in a UC-HBCU summer research program and providing courses and mentorship within an NSF IGERT program in Biophotonics.
This goal of this project is to fundamentally improve optical imaging of tissues by delivering new insights in counteracting the effects of light scattering in tissues using a model-based approach to adaptive optics that will improve image quality and penetration depth. This approach addresses many of the limitations encountered by current technologies, including: a) constraint to superficial layers due to light scattering, b) non-unique solutions and artifacts when deeper layers are accessed via wavefront shaping based on empirically based optimization schemes, c) lack of beacon sensors that could provide information to generate a compensating phase pattern, d) production of distorted focal volumes, e) iterative optimization methods that make imaging slow and, most important, f) absence of methods to model wave-based light propagation in tissue materials of meaningful volumes, leaving no model-based support for adaptive optics in tissue imaging. The new framework will be used to compute the adaptive optics signal corrections necessary to fully correct for scattering induced distortions relevant to laser scanning microscopy, and nonlinear optical microscopy in particular. The Research Plan is organized under three objectives. OBJECTIVE 1 is to develop a Virtual Microscopy Simulator (VMS) to model focused beam propagation, (linear and nonlinear) signal generation, and signal detection in turbid tissues. The input module will allow users to provide input data such as microscopy type, lens data, incident beam parameters, scattering data, nonlinear susceptibility data, detector specifications and adaptive optics data (Deformable Mirrors or Spatial Light Modulators settings). The input data will be fed into the computational engine to execute the computations. The computational module will be designed to handle all computations: a) Signal Generation, b) Signal Emission and Detection, and c) Adaptive Optics Computations. The Output module will consist of text data files of 3D focal volume data, far-field radiation data, and image data. OBJECTIVE 2 is to perform experimental validation of the VMS in tissue-mimicking phantom systems with well-defined scattering and signal generation elements. Four classes of phantoms with nonlinear optical targets will be prepared to validate the VMS: a) agarose or Sylgard only (Non-scattering), b) agarose or Sylgard with microspheres, c) Collagen Hydrogel with and w/out microspheres, and d) hybrid specimens with a layer of agarose or Sylgard matrix and a layer of collagen hydrogel with microspheres. OBJECTIVE 3 is to establish model-based adaptive optics design principles through usage of the VMS to predict the input wavefronts necessary to counteract the dispersive effects of scattering media that impede diffraction-limited focal volume formation and signal generation in scattering. The aims of the objective are to a) examine potential differences between the corrected focal volume and a theoretical, undistorted volume, b) study the performance of the algorithms from the perspective of the ideal wavefront needed to correct the image, and c) develop an adaptive algorithm that reproduces the undistorted image, even in the presence of scatterers situated between the focal volume and the collecting lens.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
