High Level Abstractions for Improving Parallel Image Reconstruction Algorithms
IWANN '09 Proceedings of the 10th International Work-Conference on Artificial Neural Networks: Part II: Distributed Computing, Artificial Intelligence, Bioinformatics, Soft Computing, and Ambient Assisted Living
| Hi-index | 0.00 |
Series expansion reconstruction methods using smooth basis functions are evaluated in the framework of electron tomography of complex biological specimens, with a special emphasis upon the computational perspective. Efficient iterative algorithms that are characterized by a fast convergence rate have been used to tackle the image reconstruction problem. The use of smooth basis functions provides the reconstruction algorithms with an implicit regularization mechanism, very appropriate for noisy conditions. High Performance Computing (HPC) techniques have been applied so as to face the computational requirements demanded by the reconstruction of large volumes. An efficient domain decomposition scheme has been devised that leads to a parallel approach with capabilities of interprocessor communication latency hiding. Comparisons with Weighted BackProjection (WBP), the standard method in the field, are presented in terms of computational demands as well as in reconstruction quality under highly noisy conditions. The combination of efficient iterative algorithms and HPC techniques have proved to be well suited for the reconstruction of large biological specimens in electron tomography, yielding solutions in reasonable computational times.