Mapping Iterative Medical Imaging Algorithm on Cell Accelerator
Algebraic reconstruction techniques require about half the number of projections as that of Fourier backprojection methods, which makes these methods safer in terms of required radiation dose. Algebraic reconstruction technique (ART) and its variant OS-SART (ordered subset simultaneous ART) are tech...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2011-01-01
|
| Series: | International Journal of Biomedical Imaging |
| Online Access: | http://dx.doi.org/10.1155/2011/843924 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832556492864094208 |
|---|---|
| author | Meilian Xu Parimala Thulasiraman |
| author_facet | Meilian Xu Parimala Thulasiraman |
| author_sort | Meilian Xu |
| collection | DOAJ |
| description | Algebraic reconstruction techniques require about half the number of projections as that of Fourier backprojection methods, which makes these methods safer in terms of required radiation dose. Algebraic reconstruction technique (ART) and its variant OS-SART (ordered subset simultaneous ART) are techniques that provide faster convergence with comparatively good image quality. However, the prohibitively long processing time of these techniques prevents their adoption in commercial CT machines. Parallel computing is one solution to this problem. With the advent of heterogeneous multicore
architectures that exploit data parallel applications, medical imaging algorithms such as OS-SART can be studied to produce increased performance. In this paper, we map OS-SART on cell broadband engine (Cell BE). We effectively use the architectural features of Cell BE to provide an efficient mapping. The Cell BE consists of one powerPC processor element (PPE) and eight SIMD coprocessors known as synergetic processor elements (SPEs). The limited memory storage on each of the SPEs makes the mapping challenging. Therefore, we present optimization techniques to efficiently map the algorithm on the Cell BE for improved performance over CPU version. We compare the performance of our proposed algorithm on Cell BE to that of Sun Fire ×4600, a shared memory machine. The Cell BE is five times faster than AMD Opteron dual-core processor. The speedup of the algorithm on Cell BE increases with the increase in the number of SPEs. We also experiment with various parameters, such as number of subsets, number of processing elements, and number of DMA transfers between main memory and local memory, that impact the performance of the algorithm. |
| format | Article |
| id | doaj-art-b2badcbd34d74b599c58f34a37d0fa9f |
| institution | Kabale University |
| issn | 1687-4188 1687-4196 |
| language | English |
| publishDate | 2011-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | International Journal of Biomedical Imaging |
| spelling | doaj-art-b2badcbd34d74b599c58f34a37d0fa9f2025-02-03T05:45:13ZengWileyInternational Journal of Biomedical Imaging1687-41881687-41962011-01-01201110.1155/2011/843924843924Mapping Iterative Medical Imaging Algorithm on Cell AcceleratorMeilian Xu0Parimala Thulasiraman1Department of Computer Science, University of Manitoba, Winnipeg, MB, R3T 2N2, CanadaDepartment of Computer Science, University of Manitoba, Winnipeg, MB, R3T 2N2, CanadaAlgebraic reconstruction techniques require about half the number of projections as that of Fourier backprojection methods, which makes these methods safer in terms of required radiation dose. Algebraic reconstruction technique (ART) and its variant OS-SART (ordered subset simultaneous ART) are techniques that provide faster convergence with comparatively good image quality. However, the prohibitively long processing time of these techniques prevents their adoption in commercial CT machines. Parallel computing is one solution to this problem. With the advent of heterogeneous multicore architectures that exploit data parallel applications, medical imaging algorithms such as OS-SART can be studied to produce increased performance. In this paper, we map OS-SART on cell broadband engine (Cell BE). We effectively use the architectural features of Cell BE to provide an efficient mapping. The Cell BE consists of one powerPC processor element (PPE) and eight SIMD coprocessors known as synergetic processor elements (SPEs). The limited memory storage on each of the SPEs makes the mapping challenging. Therefore, we present optimization techniques to efficiently map the algorithm on the Cell BE for improved performance over CPU version. We compare the performance of our proposed algorithm on Cell BE to that of Sun Fire ×4600, a shared memory machine. The Cell BE is five times faster than AMD Opteron dual-core processor. The speedup of the algorithm on Cell BE increases with the increase in the number of SPEs. We also experiment with various parameters, such as number of subsets, number of processing elements, and number of DMA transfers between main memory and local memory, that impact the performance of the algorithm.http://dx.doi.org/10.1155/2011/843924 |
| spellingShingle | Meilian Xu Parimala Thulasiraman Mapping Iterative Medical Imaging Algorithm on Cell Accelerator International Journal of Biomedical Imaging |
| title | Mapping Iterative Medical Imaging Algorithm on Cell Accelerator |
| title_full | Mapping Iterative Medical Imaging Algorithm on Cell Accelerator |
| title_fullStr | Mapping Iterative Medical Imaging Algorithm on Cell Accelerator |
| title_full_unstemmed | Mapping Iterative Medical Imaging Algorithm on Cell Accelerator |
| title_short | Mapping Iterative Medical Imaging Algorithm on Cell Accelerator |
| title_sort | mapping iterative medical imaging algorithm on cell accelerator |
| url | http://dx.doi.org/10.1155/2011/843924 |
| work_keys_str_mv | AT meilianxu mappingiterativemedicalimagingalgorithmoncellaccelerator AT parimalathulasiraman mappingiterativemedicalimagingalgorithmoncellaccelerator |