Study Of Heterogeneous Tissue Phantoms' Optical Character With Multi-spectral Reflectance Imaging

Biological tissue is a kind of turbid medium in which visible and near-infrared light is scattered strongly. Accurate determination of the optical parameters and spectral data of turbid media have great potential in bio-medical diagnosis. Determination of these parameters through the elastic scattered signal has attracted intensive research efforts in the field of optics in recent decades. The research promotes the establishment of a solid foundation for the development of new precision instruments for studies of biological tissues.The research work consists of the following two parts in the development of a multi-spectral reflectance imaging method for studying the heterogeneous tissue phantoms' optical characters:1. With a constant incident light, several reflectance images of the phantom and diffused reflectance standard was collected under different exposure times. The relationship between the exposure time t and the reflected light intensity which is the mean pixel gray value of the windowed region, and the reflected light intensities was normalized with the pixel value of the maximum exposure time and plotted as a function of t. i.e., r(t), which was used as calibration curve for calibrating the reflected light intensity against exposure time. We found that r(t) is dependent on the CCD camera only independent on the material and the incident light which makes it possible to obtain reflection images of high dynamic range.2. With the full-field illumination between 500 nm~940 nm, we captured the reflection images of 5 different heterogeneous turbid media phantoms using the same CCD camera. Then we calculated the high dynamic range reflection images from the captured images, and got the reflectance images through normalization. At the same time, we used the diffused reflectance standard to get the intensity distribution of the incident light beams. Based on the radiation transfer theory and Fresnel equations, we have developed a parallel computing software of Monte Carlo simulation, named as i MC to simulate the energy distribution in heterogeneous turbid media. We simulated the reflectance image under the same incident intensity distribution with different optical parameters of embedded materials, analyzed correlation of reflectance image contrast with the experimental and simulation parameters, such as the wave length of incident light, thickness of the embedded materials, absorption coefficient, scattering coefficient and anisotropy factor. Make a comparison between the result of experimental and simulation, we trust that we can determine the optical characteristics and geometric parameters of heterogeneous turbid media combine the multi-spectral reflectance imaging method and Monte Carlo inverse algorithm.In the end of the thesis, we summarize the process and results of the study, analysis the strengths and weaknesses of the multi-spectral reflectance imaging method. Then, we make a few suggestions to improve this method. Furth more, we make a prospect of this method in future applications.