几种典型迭代算法在生物发光断层成像中的对比研究及评估（英文）

A Comparative Study and Evaluation on Several Typical Iterative Methods for Bioluminescence Tomography

生物发光断层成像是一种新型光学分子影像技术。基于稀疏正则化的迭代算法在解决重建中病态问题起着关键的作用。将4种典型的基于l1正则化的迭代算法(内点法、同伦算法、一阶方法和拉格朗日算法)应用于重建过程中,分别进行数学推导,并从重建精度和重建速度方面进行了实验对比和性能评估。实验结果表明尽管4种方法均能较好地重建出光源位置,但时间代价和重建能量大小上存在差异,据此结果为不同情况下重建算法的选取提出建议。

Bioluminescence tomography （BLT） has recently emerged as a promising preclinical imaging modality. Iterative methods based on sparse regularization play a critical role in solving the ill-posed BLT inverse problem. Four kinds of typical iterative methods based on ll regularization were briefly introduced and applied to reconstruct the bioluminescent source location and intensity, which include interior-point methods, homotopy methods, firstorder methods, and augmented Lagrangian methods. Numerical experiments on a digital inhomogeneous mouse model and in vivo experiments were conducted to evaluate the performance of these methods in terms of localization accuracy, reconstructed intensity and power.