面向近红外脑功能成像的光学自导引漫射光断层成像方法

Optical Self-guiding Diffuse Optical Tomography towards Near-infrared Functional Brain Imaging

为降低近红外脑功能漫射光断层成像(DOT)固有的逆问题病态性,并避免多模态方法的图像配准等问题,提出了基于光学自导引提供先验功能信息的脑功能DOT方法(OT-DOT),并发展了图像重构方法.模拟验证表明:上皮厚度(TLT)已知时,OT-DOT获得的重构量化度(QR)约为传统DOT的4.2倍;当TLT的估计误差小于±10%时,OT-DOT重构的QR值可达92%以上,远远优于传统DOT;噪声鲁棒性测试表明,OT-DOT与传统DOT的噪声鲁棒性相近.利用连续光DOT测量系统的仿体实验重构结果表明,所发展的OT-DOT算法获得的重构结果优于传统DOT算法.

The image reconstruction of diffuse optical tomography （DOT） for near-infrared brain functional imaging is a severely ill-posed inverse problem. Different from the widely adopted multi-mode imaging method for providing anatomical-prior to DOT, this paper proposed an optical self-guiding scheme, called OT-DOT, and the corresponding image reconstruciton algorithm is developed. The numerical and phantom validations are performed for the proposed method. Numerical reconstruction results indicate that quantitativeness ratio （QR） reconstructed with OT-DOT for the target is about 4.2 times bigger than that with traditional DOT for the known top layer thickness（TLT）. QR from OT-DOT can be more 92% when the estimation error of TLT is less than ± 10%. The noise-robustness testing indicates the reconstruction ability of OT-DOT and traditional DOT are roughly the same. Phantom experiments performed using a continuous-wave DOT system show that the reconstructed results with the proposed method is better than those with traditional DOT.