漫射光学断层成像系统的研究与设计

Development and Design of Diffusion Optical Tomography System

目的设计并构建一套漫射光学断层成像系统。方法利用近红外激光器作为光源,仅用一个光电倍增管作为检测器。一个光学多路复用器用于切换检测通道。系统的32个通道(16个发射通道,16个检测通道)工作在连续密度波模式,用来获取256个边界检测数据。结果基于所提出的成像系统进行了实验。用intralipid作为仿体,碳素墨水作为吸收体。吸收体在浸入仿体前后分别在边界上得到一组检测数据。将这两组数据归一化后用来重建吸收系数图像。重建结果反映了吸收体的实际位置和大小。结论所提出的成像平台能够有效地用于组织光学参数的成像。但由于逆问题的欠定性很严重,重建算法中也没有考虑噪声的影响,导致重建图像的分辨率还不够高。下一代系统可以考虑利用更多的先验知识和提高系统性能来改善重建结果。

Objective To design and establish a set of diffusion optical tomography （DOT） system, Methods Near-infrared （NIR） laser was employed as the light source and only one photomultiplier tube （PMT） as the detector. An optical multiplexer was used to alter the detector channels, The 32 channels of the system, which are consisted of 16 launch channels and 16 detector channels, worked under the continuous-wave （CW） model and were used to acquire 256 boundary data. Resuits Experiments were performed based on the proposed imaging system. The intralipid was used as the tissue-like medium and the India ink as the absorber. Two sets of data on the boundary were sampled, respectively, before and after the absorber was embedded inside the tissue-like medium. The two sets of data were normalized and then used to reconstruct the absorption coefficient distribution. The recovered image reflected the real location and size of the absorber. Conclusion The proposed imaging platform can image the tissue optical parameters effectively. However the resolution of the reconstructed image was not high because the inverse problem was gravely underdetermined and the noise was not considered in the reconstruction algorithm. The recovered result in the next generation system could be improved by making more use of prior information and enhancing the performance of the system.