A new reconstruction technique used in Diffuse Optical Tomography System

Abstract

Several methods are used in the early diagnosis of breast cancer. Due to the disadvantages of these methods, the diffuse optical tomography (DOT) system is one of new methods have been developed to support these methods. In the DOT, the reconstruction technique used to create image is very important. Each reconstruction technique has some disadvantages. Thus, it is important to create an image by using the appropriate technique for DOT system. In this study, we have combined two algorithms called Truncated Conjugate Gradient (TCG) and Transpose Free Quasi Minimal Residual (TFQMR) to create a new algorithm named TCG-TFQMR. The reason of combining the two algorithms is low resolution of TCG and slow running of TFQMR. The new algorithm is first tested using simulation data created by Matlab. Then, in-vitro experiments were performed by acquiring DOT data from a breast phantom. The both reconstruction algorithms TCG and TCG-TFQMR were used to reconstruct the tissue phantom. It has been shown that TCG-TFQMR algorithm has better resolution than TCG.