Multimode Analysis of SHG Signal from Complex Biological Systems: Parameterization of Features Using Nearest-Neighbor Analysis and Wavelet Transforms

We have developed a novel computational approach for quantifying structural disorder in biomolecular lattices with nonlinear susceptibility based on analysis of polarization-modulated second harmonic signal. Transient, regional disorder at the level of molecular organization is identified using a novel signal processing algorithm sufficiently compact for near real-time analysis with a desktop computer. Global disorder within the biostructure is assessed using a two-dimensional wavelet transform of the magnitude and phase of the second harmonic signal. Selection of coefficients and the specific wavelet family~is based on topological considerations. Experimental results suggest our signal processing method represents a robust, scaleable tool that allows us to detect both regional and global alterations in signal characteristics of biostructures with a high degree of discrimination.