Automatic differentiation for GPU-accelerated 2D/3D registration
File(s)
Accepted version
Author(s)
Grabner, M
Pock, T
Gross, T
Kainz, B
Type
Journal Article
Abstract
A common task in medical image analysis is the alignment of data from different sources, e.g., X-ray images and computed tomography (CT) data. Such a task is generally known as registration. We demonstrate the applicability of automatic differentiation (AD) techniques to a class of 2D/3D registration problems which are highly computationally intensive and can therefore greatly benefit from a parallel implementation on recent graphics processing units (GPUs). However, being designed for graphics applications, GPUs have some restrictions which conflict with requirements for reverse mode AD, in particular for taping and TBR analysis. We discuss design and implementation issues in the presence of such restrictions on the target platform and present a method which can register a CT volume data set (512 × 512 × 288 voxels) with three X-ray images (512 × 512 pixels each) in 20 seconds on a GeForce 8800GTX graphics card.
Date Issued
2008-01-01
Date Acceptance
2008-01-01
Citation
Lecture Notes in Computational Science and Engineering, 2008, 64 (Advances in Automatic Differentiation), pp.259-269
ISSN
1439-7358
Publisher
Springer Verlag (Germany)
Start Page
259
End Page
269
Journal / Book Title
Lecture Notes in Computational Science and Engineering
Volume
64
Issue
Advances in Automatic Differentiation
Copyright Statement
The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-540-68942-3_23
Subjects
Optimization
medical image analysis
2D/3D registration
graphics processing unit
Publication Status
Published