An OpenCL implementation of the Gaussian pyramid and the resampler
logo

Please use this identifier to cite or link to this publication: http://hdl.handle.net/10380/3393
Nonrigid image registration is an important, but resource demanding and time-consuming task in medical image analysis. This limits its application in time-critical clinical routines. In this report we explore acceleration of two time-consuming parts of a registration algorithm by means of parallel processing using the GPU. We built upon the OpenCL-based GPU image processing framework of the recent ITK4 release, and implemented Gaussian multi-resolution strategies and a general resampling framework. We evaluated the performance gain on two multi-core machines with NVidia GPUs, and compared to an existing ITK4 CPU implementation. A speedup factor of ~2-4 was realized for the multi-resolution strategies and a speedup factor of ~10-46 was achieved for resampling, for larger images (~10^8 voxels).
Data
minus 1 Dataset (3Mb)
Code
plus Automatic Testing Results by Insight-Journal Dashboard on Wed Mar 6 16:16:31 2013 for revision #2
starstarstarstarstar expertise: 5 sensitivity: 5
plus Automatic Testing Results by Insight-Journal Dashboard on Tue Jan 29 17:14:32 2013 for revision #1
starstarstarstarstar expertise: 5 sensitivity: 5

Reviews
There is no review at this time. Be the first to review this publication!

Quick Comments
Comment by Denis Shamonin yellow
1. The source code has been merged with latest elastix see https://github.com/SuperElastix/elastix
2. The source code has been tested with latest ITK
ITK 4.12.2 (25 Oct 2017)
CUDA/v8.0 (NVidia driver 378.49)
Intel OpenCL SDK (version 7.0.0.2519) on CPU
3. The source code builds and tests performed using elastix dashboard:
http://my.cdash.org/index.php?project=elastix
4. The full paper:
http://journal.frontiersin.org/article/10.3389/fninf.2013.00050/full
5. The latest GPU performance results could be found in section supplemental data of the full paper:
http://journal.frontiersin.org/File/DownloadFile/65120/octet-stream/Presentation%201.PDF/12/1/66699
6. The recent GPU/CPU performance experiments could be found in directory test_logs see source code.


Resources
backyellow
Download All
Download Paper , View Paper
Download Source code

Statistics more
backyellow
Global rating: starstarstarstarstar
Review rating: starstarstarstarstar [review]
Code rating: starstarstarstarstar
Paper Quality: plus minus
1 comment

Information more
backyellow
Categories: Image pyramids, Resampling, Transforms
Keywords: image registration, parallelization, GPU, OpenCL, GPUResampleImageFilter, GenericMultiResolutionPyramidImageFilter, elastix
Tracking Number: NWO NRG-2010.02, NWO 639.021.124
Toolkits: ITK, CMake
Export citation:

Share
backyellow
Share

Linked Publications more
backyellow
Importing Contours from DICOM-RT Structure Sets Importing Contours from DICOM-RT Structure Sets
by Dowling J., Malaterre M., Greer P.B., Salvado O.
A Skull-Stripping Filter for ITK A Skull-Stripping Filter for ITK
by Bauer S., Fejes T., Reyes M.

View license
Loading license...

Send a message to the author
main_flat
ISSN 2327-770X
Powered by Midas