BSIRT: A Block-Iterative SIRT Parallel Algorithm Using Curvilinear Projection Model

Large-field high-resolution electron tomography enables visualizing detailed mechanisms under global structure. As field enlarges, the distortions of reconstruction and processing time become more critical. Using the curvilinear projection model can improve the quality of large-field ET reconstruction, but its computational complexity further exacerbates the processing time. Moreover, there is no parallel strategy on GPU for iterative reconstruction method with curvilinear projection. Here we propose a new Block-iterative SIRT parallel algorithm with the curvilinear projection model (BSIRT) for large-field ET reconstruction, to improve the quality of reconstruction and accelerate the reconstruction process. We also develop some key techniques, including block-iterative method with the curvilinear projection, a scope-based data decomposition method and a page-based data transfer scheme to implement the parallelization of BSIRT on GPU platform. Experimental results show that BSIRT can improve the reconstruction quality as well as the speed of the reconstruction process.

[1]  W. Marsden.  I and J , 2012 .

[2]  José María Carazo,et al.  Parallelization and comparison of 3D iterative reconstruction algorithms , 2004, 12th Euromicro Conference on Parallel, Distributed and Network-Based Processing, 2004. Proceedings..

[3]  Reiner Hegerl,et al.  Cryo-electron tomography of biological specimens , 2006, IEEE Signal Processing Magazine.

[4]  Gabor T. Herman,et al.  Algebraic reconstruction techniques can be made computationally efficient [positron emission tomography application] , 1993, IEEE Trans. Medical Imaging.

[5]  Kai Zhang,et al.  Three-dimensional reconstruction using an adaptive simultaneous algebraic reconstruction technique in electron tomography. , 2011, Journal of Structural Biology.

[6]  Kevin Barraclough,et al.  I and i , 2001, BMJ : British Medical Journal.

[7]  Renmin Han,et al.  Iterative Methods in Large Field Electron Microscope Tomography , 2013, SIAM J. Sci. Comput..

[8]  Y. Censor,et al.  Parallel Optimization: Theory, Algorithms, and Applications , 1997 .

[9]  Y. Censor,et al.  Parallel Optimization:theory , 1997 .

[10]  Rupa G. Mehta,et al.  Impact of Outlier Removal and Normalization Approach in Modified k-Means Clustering Algorithm , 2011 .

[11]  José-Jesús Fernández,et al.  Evaluation of a Multicore-Optimized Implementation for Tomographic Reconstruction , 2012, PloS one.

[12]  Albert F. Lawrence,et al.  Parallel Processing and Large-Field Electron Microscope Tomography , 2009, 2009 WRI World Congress on Computer Science and Information Engineering.

[13]  Y. Censor,et al.  BICAV: a block-iterative parallel algorithm for sparse systems with pixel-related weighting , 2001, IEEE Transactions on Medical Imaging.

[14]  M Cao,et al.  Formation and reduction of streak artefacts in electron tomography , 2010, Journal of microscopy.

[15]  Mark H Ellisman,et al.  Transform-based backprojection for volume reconstruction of large format electron microscope tilt series. , 2006, Journal of structural biology.

[16]  Inmaculada García,et al.  Electron tomography of complex biological specimens on the Grid , 2007, Future Gener. Comput. Syst..

[17]  Sébastien Phan,et al.  TxBR montage reconstruction for large field electron tomography. , 2012, Journal of structural biology.

[18]  Yair Censor,et al.  BICAV: An Inherently Parallel Algorithm for Sparse Systems with Pixel-Dependent Weighting , 2001, IEEE Trans. Medical Imaging.