Bioluminescence tomography using eigenvectors expansion and iterative solution for the optimized permissible source region.
Bioluminescence tomography using eigenvectors expansion and iterative solution for the optimized permissible source region.
复制标题
DOI：
10.1364/boe.2.003179
复制
发表时间：
2011-11-01
影响因子：
3.4
通讯作者：
Patterson MS
中科院分区：
文献类型：
--
作者：
Naser MA;Patterson MS
提出了一种生物发光层析成像(BLT)重建算法。该算法利用扩散方程和有限元方法对不同波长的格林函数进行了数值计算。利用漫反射光学层析成像(DOT)重建用于计算格林函数的光学性质，并假设解剖信息由X射线计算机层析成像或其他方法提供。利用格林函数和测量的光通量率组成一个对称的方程组，并求解由此产生的本征值问题，得到该对称方程组的本征向量。可以从所获得的特征向量形成空间，并且重构的信源被写为对应于非零特征值的特征向量的展开。求出展开系数，得到重建的BL源分布。该问题通过使用允许的源区迭代地解决，该允许的源区通过移除对源有贡献的低概率节点来缩小。在整个过程中，允许区域从整个对象缩小到只有几个节点。对重建光源的最佳估计被选择为使计算的光通量速率和测量的光通量速率之间的差异最小化。三维模拟结果表明，无论是局域多个震源还是大范围非均匀震源，重建的震源在位置、震级、大小和总功率方面都与实际震源吻合较好。
A reconstruction algorithm for bioluminescence tomography (BLT) has been developed. The algorithm numerically calculates the Green’s function at different wavelengths using the diffusion equation and finite element method. The optical properties used in calculating the Green’s function are reconstructed using diffuse optical tomography (DOT) and assuming anatomical information is provided by x-ray computed tomography or other methods. A symmetric system of equations is formed using the Green’s function and the measured light fluence rate and the resulting eigenvalue problem is solved to get the eigenvectors of this symmetric system of equations. A space can be formed from the eigenvectors obtained and the reconstructed source is written as an expansion of the eigenvectors corresponding to non-zero eigenvalues. The coefficients of the expansion are found to obtain the reconstructed BL source distribution. The problem is solved iteratively by using a permissible source region that is shrunk by removing nodes with low probability to contribute to the source. Throughout this process the permissible region shrinks from the entire object to just a few nodes. The best estimate of the reconstructed source is chosen that which minimizes the difference between the calculated and measured light fluence rates. 3D simulations presented here show that the reconstructed source is in good agreement with the actual source in terms of locations, magnitudes, sizes, and total powers for both localized multiple sources and large inhomogeneous source distributions.