| The light microscope has been designed and optimized for the study of very thin specimens or tissue sections. A recent trend in biology, however, has been to study intact and often living three-dimensional specimens using the light microscope. In order to optimize methods of performing three-dimensional microscopy on intact specimens, such as Confocal Laser Scanning microscopy and Optical Serial Sectioning microscopy, it is important that the three-dimensional imaging properties of these systems be well understood. A number of theoretical models exist of three-dimensional image formation in the light microscope. However, these are based on an ideal, aberration-free, diffraction-limited lens system and it has been observed in practice that there are significant differences between measured point-spread functions and the predictions of these models. Many of these differences can be accounted for by aberrations in the microscope system that are introduced when the microscope is used under non-design conditions. Thus, an accurate theoretical model of the imaging system's three-dimensional point-spread function must incorporate these aberrations.;In this dissertation, a mathematical expression is developed, based on both geometrical and physical optics, of the aberration introduced when an immersion objective lens is used under non-design conditions. A non-design condition exists when any of the layers that separate the front element of the objective lens from the specimen plane of interest have a thickness or refractive index which differs from that of the design system. The relationships between the resultant aberration expression and the three-dimensional pointspread functions of both Confocal Laser Scanning microscopy and Optical Serial Sectioning microscopy are determined and are used to simulate three-dimensional images formed by the microscope for point source objects under a variety of optical conditions. Experiments that were used to test the accuracy of the model predictions are described and their results are presented. The effects of object space and image space parameters on the amount of aberration present in the system as well as the use of optical correction of aberrations, the selection of an objective lens to minimize aberrations and the use of numerical correction in processing Optical Serial Sectioning microscopy data are also discussed. |
