Paper Abstract and Keywords |
Presentation |
2015-03-03 13:40
Evaluation of accuracy and processing time of GPU-based 3D non-linear registration Haruhiko Futada (MHIR), Takeshi Nagata (MHIR/Tsukuba Univ.), Hidemasa Maekawa (MHIR), Takumi Tanikawa (Asahikawa Med. Univ.), Koichi Ishizu (Kyoto Univ.), Masami Mukai (NIRS), Koji Uemura (Kagawa Univ.), Tadashi Kamada (NIRS), Masayoshi Seki (Global For), Yutaka Ando (NIRS) MI2014-106 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
When a patient is treated with multiple radiotherapy, radiation treatment planning that doesn't overlap with the previous radiotherapy is desired. To calculate sum of multiple dose distributions, we have worked on 3D non-linear registration of same patient CT images taken at differrent times. The processing time is one of the tasks. In this study, 3D non-linear registraion method that is suited for parallel processing is newly implemented on GPU for reducing processing time. The registration accuracy and the processing time are evaluated. Results showed that the GPU implementation is 16 times faster than CPU implementation with same accuracy. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Radiation Treatment planning / Non-linear registration / GPU / GPGPU / / / / |
Reference Info. |
IEICE Tech. Rep., vol. 114, no. 482, MI2014-106, pp. 259-263, March 2015. |
Paper # |
MI2014-106 |
Date of Issue |
2015-02-23 (MI) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
MI2014-106 |
|