基于非均匀采样数据的SD-OCT成像算法研究  

ON IMAGING ALGORITHM FOR SPECTRAL DOMAIN OPTICAL COHERENCE TOMOGRAPHY USING NONUNIFORM SAMPLING DATA

作　　者：刘玉喜[1] 修晓玉 周国辉[1] 

机构地区：[1]哈尔滨师范大学计算机科学与信息工程学院,黑龙江哈尔滨150025

出　　处：《计算机应用与软件》2014年第4期235-238,共4页Computer Applications and Software

基　　金：哈尔滨市智能教育与信息工程重点实验室基金项目(081203)

摘　　要：在频域光学相干层析成像(SD-OCT)中,数据通常通过波数域的非均匀采样得到。在进行快速傅里叶变换得到图像之前需要进行一个数据重采样过程。数据重采样会造成额外的计算负担,而且会带来一定的数据误差。采用一种逆成像算法得到SDOCT图像,将SD-OCT系统建模为一系列线性方程组,通过求解一个逆问题以实现SD-OCT系统的成像。利用全变差(TV)作为限制条件以保留图像的边缘信息,然后由SD-OCT测量数据直接估计样本的二维互相关。仿真结果表明,和传统方法相比,该算法所得到的噪声残余量更低。同时还验证了利用TV限制条件以抑制SD-OCT中对衰落的敏感性的可能性。In spectral-domain optical coherence tomography（ SD-OCT）,the data are usually collected by nonuniform sampling in wave number domain. There has the need of data re-sampling process before the conventional fast Fourier transform being applied to reconstruct an image. Data re-sampling will cause extra computation burden and often introduces certain errors to data. Instead,we develop an inverse imaging approach to reconstruct an SD-OCT image,model the SD-OCT as a series of linear equations,and implement SD-OCT system imaging by finding the solution of the inverse problem. We make use of total variation（ TV） as a constraint to preserve image＇s edge information,and estimate the two-dimensional cross-correlation of a sample directly from SD-OCT measurements. Simulation results indicate that compared with conventional method,our technique gives a smaller noise residual. The potential of using the TV constraint to suppress the sensitivity falloff in SD-OCT is also demonstrated with experiment data in the paper.

关 键 词：SD-OCT 非均匀采样 全变差 重采样 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]