Skip to Main Content
Different skin and sub-surface tissues have distinct or unique reflectance pattern which help us differentiate normal and cancerous tissues. Optical means of characterizing tissues have gained importance due to its noninvasive nature. Spectral characteristics of the components provide useful information to identify the components, because different chromophores have different spectroscopic responses to electromagnetic waves of a certain energy band. Even though different tissues have different reflectance spectra, visually they all look alike. The large volumes of multidimensional data obtained are analyzed using a principal component analysis (PCA) method to quickly and objectively extract any useful signature associated with a skin disorder under study. An optical fiber spectrometer is set up for collection of diffuse reflectance data from different skin conditions. The method involves exposure of skin surface to white light produced by an incandescent source. These back scattered photons emerging from various layers of tissue are detected by spectrometer resulting in diffuse reflectance data. For the present study different skin conditions like warts, vitiligo, thrombus (due to injury) and angioma are chosen. The spectral data obtained from the scan are plotted and compared. More or less the shapes of the spectral curves for various skin conditions resemble. In order to characterize and differentiate different diseased tissue principal component analysis method is carried out. The preliminary results show that the proposed principal component analysis method is able to enhance the peculiar characteristics of the diseased diffuse reflectance spectra.