Abstract

FT-IR microspectroscopy in combination with digital image reassembling methodologies have been used to characterize distinct tissue structures within thin sections from the human prostate. The spatially resolved microspectroscopic data were collected with a resolution near the diffraction limit (about 8 μm) by using a HgCdTe focal plane array detector-based infrared imaging instrument. While IR imaging based on distinct spectral parameters such as intensities, frequency values, or half-widths (the univariate imaging technique of chemical mapping) often gives unsatisfactory results, multivariate data analysis techniques (e.g. hierarchical clustering or principal component analysis) confirmed standard histopathological techniques and turned out to be helpful to discriminate reliably between different tissues structures.

© 2001 OSA/SPIE