Research On Optical Tomography Of Mammary Gland With Near-Infrared

Breast cancer is one of the most common diseases among females, and recently theincidence of breast cancer has been on the rise. Technology of medical imaging plays anindispensable role in the detection and diagnosis of breast cancer. Near-Infrared, with itsnon-invasive, non-radiative traits, has been applied to medical fields, such as breast cancerdetection, functional imaging of blood oxygen and neonatal brain monitoring.We conduct our research study on NIR imaging and optical tomography withchemical imitation of mammary gland, including four sections. In the first place, we diveourselves in optical theory of biological tissue and explore Finite Element Method andMonte Carlo simulation to set up forward model of photon propagation, then we deduceand apply Levenberg-Marguardt algorithm to solve least squares form of inverse problem.Secondly, in the hardware design part, we give a detailed description of the principle andimplementation of our NIR motion and scanning control device. Then, we come up withthe simulation of forward model with chemical imitation of mammary gland and theinverse reconstruction with nonlinear optimization theory. Finally, we demonstrate severalexperimental results to validate our NIR tomographic reconstruction method.We carry out our research with chemical imitation of mammary gland, and measurethe Gaussian scattering feature to simulate photon migration of forward model. Throughthe imaging system which has single light source, dual optical path, we obtain both sidesof transilluminated image of breast tissues, and construct least squares form of inversemodel with theoretical prediction data and measured data. Using quadprog algorithmwhich excels in multi-parameter optimization, we solve quadratic programming form ofobjective function iteratively. Experiments show that proposed algorithm can reconstructtomographic image of breast tissue to a certain extend.