GPU-accelerated block matching algorithm for deformable registration of lung CT images

Min Li, Zhikang Xiang, Liang Xiao, Edward Castillo, Richard Castillo, Thomas Guerrero

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Deformable registration (DR) is a key technology in the medical field. However, many of the existing DR methods are time-consuming and the registration accuracy needs to be improved, which prevents their clinical applications. In this study, we propose a parallel block matching algorithm for lung CT image registration, in which the sum of squared difference metric is modified as the cost function and the moving least squares approach is used to generate the full displacement field. The algorithm is implemented on Graphic Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Results show that the proposed parallel block matching method achieves a fast runtime while maintaining an average registration error (standard deviation) of 1.08 (0.69) mm.

Original languageEnglish (US)
Title of host publicationProceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages292-295
Number of pages4
ISBN (Electronic)9781467380867
DOIs
StatePublished - Jun 10 2016
Externally publishedYes
Event3rd IEEE International Conference on Progress in Informatics and Computing, PIC 2015 - Nanjing, China
Duration: Dec 18 2015Dec 20 2015

Other

Other3rd IEEE International Conference on Progress in Informatics and Computing, PIC 2015
CountryChina
CityNanjing
Period12/18/1512/20/15

Fingerprint

Image registration
Cost functions
Graphics processing unit

Keywords

  • block matching
  • Deformable registration
  • Graphic Processing Unit

ASJC Scopus subject areas

  • Computer Science Applications
  • Computer Vision and Pattern Recognition
  • Information Systems

Cite this

Li, M., Xiang, Z., Xiao, L., Castillo, E., Castillo, R., & Guerrero, T. (2016). GPU-accelerated block matching algorithm for deformable registration of lung CT images. In Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015 (pp. 292-295). [7489856] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/PIC.2015.7489856

GPU-accelerated block matching algorithm for deformable registration of lung CT images. / Li, Min; Xiang, Zhikang; Xiao, Liang; Castillo, Edward; Castillo, Richard; Guerrero, Thomas.

Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015. Institute of Electrical and Electronics Engineers Inc., 2016. p. 292-295 7489856.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Li, M, Xiang, Z, Xiao, L, Castillo, E, Castillo, R & Guerrero, T 2016, GPU-accelerated block matching algorithm for deformable registration of lung CT images. in Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015., 7489856, Institute of Electrical and Electronics Engineers Inc., pp. 292-295, 3rd IEEE International Conference on Progress in Informatics and Computing, PIC 2015, Nanjing, China, 12/18/15. https://doi.org/10.1109/PIC.2015.7489856
Li M, Xiang Z, Xiao L, Castillo E, Castillo R, Guerrero T. GPU-accelerated block matching algorithm for deformable registration of lung CT images. In Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015. Institute of Electrical and Electronics Engineers Inc. 2016. p. 292-295. 7489856 https://doi.org/10.1109/PIC.2015.7489856
Li, Min ; Xiang, Zhikang ; Xiao, Liang ; Castillo, Edward ; Castillo, Richard ; Guerrero, Thomas. / GPU-accelerated block matching algorithm for deformable registration of lung CT images. Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015. Institute of Electrical and Electronics Engineers Inc., 2016. pp. 292-295
@inproceedings{0309db5e88eb4a7ebd5b54332a8ba9b4,
title = "GPU-accelerated block matching algorithm for deformable registration of lung CT images",
abstract = "Deformable registration (DR) is a key technology in the medical field. However, many of the existing DR methods are time-consuming and the registration accuracy needs to be improved, which prevents their clinical applications. In this study, we propose a parallel block matching algorithm for lung CT image registration, in which the sum of squared difference metric is modified as the cost function and the moving least squares approach is used to generate the full displacement field. The algorithm is implemented on Graphic Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Results show that the proposed parallel block matching method achieves a fast runtime while maintaining an average registration error (standard deviation) of 1.08 (0.69) mm.",
keywords = "block matching, Deformable registration, Graphic Processing Unit",
author = "Min Li and Zhikang Xiang and Liang Xiao and Edward Castillo and Richard Castillo and Thomas Guerrero",
year = "2016",
month = "6",
day = "10",
doi = "10.1109/PIC.2015.7489856",
language = "English (US)",
pages = "292--295",
booktitle = "Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - GPU-accelerated block matching algorithm for deformable registration of lung CT images

AU - Li, Min

AU - Xiang, Zhikang

AU - Xiao, Liang

AU - Castillo, Edward

AU - Castillo, Richard

AU - Guerrero, Thomas

PY - 2016/6/10

Y1 - 2016/6/10

N2 - Deformable registration (DR) is a key technology in the medical field. However, many of the existing DR methods are time-consuming and the registration accuracy needs to be improved, which prevents their clinical applications. In this study, we propose a parallel block matching algorithm for lung CT image registration, in which the sum of squared difference metric is modified as the cost function and the moving least squares approach is used to generate the full displacement field. The algorithm is implemented on Graphic Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Results show that the proposed parallel block matching method achieves a fast runtime while maintaining an average registration error (standard deviation) of 1.08 (0.69) mm.

AB - Deformable registration (DR) is a key technology in the medical field. However, many of the existing DR methods are time-consuming and the registration accuracy needs to be improved, which prevents their clinical applications. In this study, we propose a parallel block matching algorithm for lung CT image registration, in which the sum of squared difference metric is modified as the cost function and the moving least squares approach is used to generate the full displacement field. The algorithm is implemented on Graphic Processing Unit (GPU) with the Compute Unified Device Architecture (CUDA). Results show that the proposed parallel block matching method achieves a fast runtime while maintaining an average registration error (standard deviation) of 1.08 (0.69) mm.

KW - block matching

KW - Deformable registration

KW - Graphic Processing Unit

UR - http://www.scopus.com/inward/record.url?scp=84979711150&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84979711150&partnerID=8YFLogxK

U2 - 10.1109/PIC.2015.7489856

DO - 10.1109/PIC.2015.7489856

M3 - Conference contribution

SP - 292

EP - 295

BT - Proceedings of 2015 IEEE International Conference on Progress in Informatics and Computing, PIC 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -