Mid infrared spectroscopic imaging and tomography

Fourier Transform Infrared (FT-IR) Spectroscopic Imaging is an optical technique that simultaneously provides morphological (two dimensional) and mid-infrared spectral (one dimensional) information for a thin specimen under observation. In this dissertation, we develop a new instrument that extends the capabilities of FT-IR imaging to provide three dimensional morphological information along with mid-infrared spectral information on the microscopic length scale. We first demonstrate that spectroscopic information from the mid-infrared can be used to perform histologic recognition and cancer detection in prostate, breast and skin cancer tissues. These studies perform tissue segmentation on FT-IR imaging data from several hundred patients. The segmentation algorithms and instrumentation used are described in detail. We perform a rigorous analysis of FT-IR imaging instruments and data so as to understand the limits and limitations of current state-ofthe- art instrumentation. We then propose a new instrument that adds a new dimension of information to current instruments and establish the new state-of-the-art in mid-infrared imaging. The theoretical basis of such instrumentation is presented in detail and techniques for visualization of four dimensional data obtained from the new instrument are outlined.