Visualization of cells and subcellular organelles are currently carried out using available microscopy methods such as
cryoelectron microscopy, and fluorescence microscopy. These methods require external labeling using fluorescent dyes
and extensive sample preparations to access the subcellular structures. However, Raman micro-spectroscopy provides a
non-invasive, label-free method for imaging the cells with chemical specificity at sub-micrometer spatial resolutions.
The scope of this paper is to image the biochemical/molecular distributions in cells associated with cancerous changes.
Raman map data sets were acquired from the human cervical carcinoma cell lines (HeLa) after fixation under 785 nm
excitation wavelength. The individual spectrum was recorded by raster-scanning the laser beam over the sample with
1μm step size and 10s exposure time. Images revealing nucleic acids, lipids and proteins (phenylalanine, amide I) were
reconstructed using univariate methods. In near future, the small pixel to pixel variations will also be imaged using
different multivariate methods (PCA, clustering (HCA, K-means, FCM)) to determine the main cellular constitutions.
The hyper-spectral image of cell was reconstructed utilizing the spectral contrast at different pixels of the cell (due to the
variation in the biochemical distribution) without using fluorescent dyes. Normal cervical squamous cells will also be
imaged in order to differentiate normal and cancer cells of cervix using the biochemical changes in different grades of
cancer. Based on the information obtained from the pseudo-color maps, constructed from the hyper-spectral cubes, the
primary cellular constituents of normal and cervical cancer cells were identified.