東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Accelerated Tomographic Image Recons...

Pan, Hui.

FindBook

Google Book

Amazon

博客來

Accelerated Tomographic Image Reconstruction of SPECT- CT Using GPU Parallelization.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Accelerated Tomographic Image Reconstruction of SPECT- CT Using GPU Parallelization./
作者:	Pan, Hui.
面頁冊數:	135 p.
附註:	Source: Dissertation Abstracts International, Volume: 76-12(E), Section: B.
Contained By:	Dissertation Abstracts International76-12B(E).
標題:	Molecular biology. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3664017
ISBN:	9781339099361

Accelerated Tomographic Image Reconstruction of SPECT- CT Using GPU Parallelization.
Pan, Hui.

Accelerated Tomographic Image Reconstruction of SPECT- CT Using GPU Parallelization. - 135 p.

Source: Dissertation Abstracts International, Volume: 76-12(E), Section: B.

Thesis (Ph.D.)--Florida Institute of Technology, 2015.

A graphics processing unit (GPU), also occasionally called a visual processing unit (VPU), is a specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display. With the increasing needs of the very active computer graphics development community, the GPU has become an integral part of today's mainstream computing systems. Especially over the past six years, GPUs have been evolving at a rapid rate. Due to the massively parallel architecture and relatively low cost, GPUs have become powerful platforms for scientific computation. For tomography, iterative reconstruction algorithms pose tremendous computational challenges due to the massive computation requirements. GPUs provide an affordable platform to these requirements. In this work, we developed some GPU enabled algorithms to make use of acceleration techniques to speed up the reconstruction processing. Single Photon Emission Computed Tomography (SPECT) can require two types of images: static and dynamic. In the static case, we parallelized the Maximum likelihood Expectation Maximization Algorithm (MLEM), Ordered-Subsets Expectation Maximization (OSEM), Computed Tomography (CT), and the Point Spread Function algorithm (PSF). In the dynamic case, we parallelized the dynamic MLEM. All the algorithms performances are validated by the same algorithms but in the CPU version. For each algorithm, as the precondition for the same reconstructed results, we compared the experiment evaluation of scalability between the CPU and GPU versions. Moreover, we reorganized the GPU thread balancing to improve the GPU algorithm performance. In addition, we developed a data organization system, which is called ReMI, to prevent data loss or corruption. Our all experiments dataset were downloaded from this system.

ISBN: 9781339099361Subjects--Topical Terms:

517296
Molecular biology.

Accelerated Tomographic Image Reconstruction of SPECT- CT Using GPU Parallelization.
LDR:02782nmm a2200289 4500 001 2115884
005 20170417071252.5
008 180830s2015 ||||||||||||||||| ||eng d
020 $a 9781339099361
035 $a (MiAaPQ)AAI3664017
035 $a AAI3664017
040 $a MiAaPQ $c MiAaPQ
100 1 $a Pan, Hui. $3 2179412
245 1 0 $a Accelerated Tomographic Image Reconstruction of SPECT- CT Using GPU Parallelization.
300 $a 135 p.
500 $a Source: Dissertation Abstracts International, Volume: 76-12(E), Section: B.
500 $a Adviser: Debasis Mitra.
502 $a Thesis (Ph.D.)--Florida Institute of Technology, 2015.
520 $a A graphics processing unit (GPU), also occasionally called a visual processing unit (VPU), is a specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display. With the increasing needs of the very active computer graphics development community, the GPU has become an integral part of today's mainstream computing systems. Especially over the past six years, GPUs have been evolving at a rapid rate. Due to the massively parallel architecture and relatively low cost, GPUs have become powerful platforms for scientific computation. For tomography, iterative reconstruction algorithms pose tremendous computational challenges due to the massive computation requirements. GPUs provide an affordable platform to these requirements. In this work, we developed some GPU enabled algorithms to make use of acceleration techniques to speed up the reconstruction processing. Single Photon Emission Computed Tomography (SPECT) can require two types of images: static and dynamic. In the static case, we parallelized the Maximum likelihood Expectation Maximization Algorithm (MLEM), Ordered-Subsets Expectation Maximization (OSEM), Computed Tomography (CT), and the Point Spread Function algorithm (PSF). In the dynamic case, we parallelized the dynamic MLEM. All the algorithms performances are validated by the same algorithms but in the CPU version. For each algorithm, as the precondition for the same reconstructed results, we compared the experiment evaluation of scalability between the CPU and GPU versions. Moreover, we reorganized the GPU thread balancing to improve the GPU algorithm performance. In addition, we developed a data organization system, which is called ReMI, to prevent data loss or corruption. Our all experiments dataset were downloaded from this system.
590 $a School code: 0473.
650 4 $a Molecular biology. $3 517296
650 4 $a Biographies. $3 795061
650 4 $a Information science. $3 554358
690 $a 0307
690 $a 0304
690 $a 0723
710 2 $a Florida Institute of Technology. $3 718970
773 0 $t Dissertation Abstracts International $g 76-12B(E).
790 $a 0473
791 $a Ph.D.
792 $a 2015
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3664017