Raman chemical imaging of Lipid and protein distribution in breast is
performed without the use of invasive contrast agents. Instead, tissue
component discrimination is based on the unique vibrational spectra i
ntrinsic to lipids and proteins. Component discrimination is possible
using ratiometric image analysis techniques. However, chemical image c
ontrast is enhanced when techniques are employed that use the entire R
aman spectrum, including peak shift analysis and multivariate image an
alysis. Visualization of breast tissue components is an essential step
in the development of a quantitative Raman 'optical biopsy' technique
suitable for the non-invasive detection and classification of breast
cancer. A high spatial/spectral resolution Raman imaging microscope is
used that employs a liquid crystal tunable filter (LCTF). The imaging
performance of the Raman microscope is quantitated with the measureme
nt of the imaging system's contrast transfer function (CTF). A compari
son between the theoretical (diffraction limited) CTF and the measured
CTF establishes that the LCTF does not produce appreciable image degr
adation. (C) 1997 by John Whey & Sons, Ltd.