首先研究了LOT中光线倾斜角度对光在组织体内分布的影响, 在此基础上设计并完成了压缩倾角薄层光学层析成像(cdaLOT)系统, 提出了矫正扫描振镜枕型畸变的新方法, 并发展了基于虚拟源-扩散近似技术和GPU的LOT快速图像重建算法.光路模拟结果表明, cdaLOT系统可使入射主光线倾角减小为传统LOT系统的1/2.cdaLOT系统将测量值与蒙特卡洛模拟结果的相对误差由传统系统的38%降低为18%, 从而缓解了实际测量方式和正向数学模型的不匹配性.仿体成像实验表明: 重建异质体的吸收系数、位置和形状与真实情况基本相符.

The influence of the dip angle on light distribution inside tissue was investigated. Based on the investigation, a LOT system with condensed dip-angle (cdaLOT) over the whole detection range was developed. A new method for correcting the deflection-angle of galvanometer mirrors was also proposed. A fast image reconstruction algorithm for LOT was also developed based on the virtual-source diffusion approximation and GPU. Light path simulation showed that the developed cdaLOT system decreased the dip angle of the incident light to 1/2 of that of the traditional LOT system. The relative deviation of the measured data to MC results was about 18% for cdaLOT while that for traditional LOT was as large as 38%. Reconstructed images showed that the reconstructed absorption coefficient, and the position and shape of the inhomogeneity coincided with the real target.