Designing 3-D Nonlinear Diffusion Filters for High Performance Cluster Computing

Bruhn, Andrés; Jakob, Tobias; Fischer, Markus; Kohlberger, Timo; Weickert, Joachim; Brüning, Ulrich; Schnörr, Christoph

doi:10.1007/3-540-45783-6_35

Cited by 7 publications

(10 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, the performance of the current architecture represents an improvement over a corresponding implementation using a 256-processor computing cluster reported previously [100]. That design, however, was based on a partial sorting technique and cannot be easily extended to kernel sizes beyond 3.…”

Section: Filtering Performancementioning

confidence: 89%

“…Accelerated implementations of 3D anisotropic diffusion filtering using computing clusters have also been reported. Bruhn et al [99,100] have reported an approach using a 256-node Myrinet cluster, whereas Tabik et al [101] have explored multiple parallel programming paradigms built on message passing and shared-memory architectures.…”

Section: Multiobjective Optimizationmentioning

confidence: 99%

“…anisotropic diffusion filtering were targeted toward multiprocessor clusters [100,101]. Despite the near-linear speedup offered by these techniques, the need to employ up to 256 processors to achieve real-time performance makes them less suitable for clinical deployment.…”

Section: Motivationmentioning

confidence: 99%

See 2 more Smart Citations

High-Performance 3D Image Processing Architectures for Image-Guided Interventions

Dandekar¹

2008

View full text Add to dashboard Cite

Minimally invasive image-guided interventions (IGIs) are time and cost efficient, minimize unintended damage to healthy tissues, and lead to faster patient recovery. Advanced three-dimensional (3D) image processing is a critical need for navigation during IGIs. However, achieving on-demand performance, as required by IGIs, for these image processing operations using software-only implementations is challenging because of the sheer size of the 3D images, and memory and compute intensive nature of the operations. This dissertation, therefore, is geared toward developing high-performance 3D image processing architectures, which will enable improved intraprocedural visualization and navigation capabilities during IGIs.In this dissertation we present an architecture for real-time implementation of 3D filtering operations that are commonly employed for preprocessing of medical images. This architecture is approximately two orders of magnitude faster than corresponding software implementations and is capable of processing 3D medical images at their acquisition speeds. Report Documentation PageCombining complementary information through registration between pre-and intraprocedural images is a fundamental need in the IGI workflow. Intensity-based deformable registration, which is completely automatic and locally accurate, is a promising approach to achieve this alignment. These algorithms, however, are extremely compute intensive, which has prevented their clinical use. We present an FPGA-based architecture for accelerated implementation of intensity-based deformable image registration. This high-performance architecture achieves over an order of magnitude speedup when compared with a corresponding software implementation and reduces the execution time of deformable registration from hours to minutes while offering comparable image registration accuracy.Furthermore, we present a framework for multiobjective optimization of finite-precision implementations of signal processing algorithms that takes into account multiple conflicting objectives such as implementation accuracy and hardware resource consumption. The evaluation that we have performed in the context of FPGA-based image registration demonstrates that such an analysis can be used to enhance automated hardware design processes, and efficiently identify a system configuration that meets given design constraints. In addition, we also outline two novel clinical applications that can directly benefit from these developments and demonstrate the feasibility of our approach in the context of these applications. These advances will ultimately enable integration of 3D image processing into clinical workflow.

show abstract

Section: Filtering Performancementioning

confidence: 89%

Section: Multiobjective Optimizationmentioning

confidence: 99%

See 1 more Smart Citation

High-Performance 3D Image Processing Architectures for Image-Guided Interventions

Dandekar¹

2008

View full text Add to dashboard Cite

show abstract

“…In our algorithm, it is the task of the weight factor k to model the behavior in high confidence points versus the one if the confidence is low. We assign to each point p in the image a confidence k(p) in the interval [0,1]. Therefore, at each step, if the confidence in a point is high, the transformation in that point will be mainly driven by the similarity term.…”

Section: Trade-off Between Matching and Regularizationmentioning

confidence: 99%

“…Bruhn [1] proposes an implementation that requires clusters connected through high performance (hence expensive) networks. At each step, the image that is being filtered is redistributed, requiring all processors to communicate to each other.…”

Section: Parallelizationmentioning

confidence: 99%

Grid Enabled Non-rigid Registration with a Dense Transformation and a priori Information

Stefanescu

Pennec

Ayache

2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Multi-subject non-rigid registration algorithms using dense transformations often encounter cases where the transformation to be estimated requires a large spatial variability. In these cases, linear regularization methods are not sufficient. In this paper, we present an algorithm that uses a priori information about the nature of the images in order to find more adapted deformations. We also present a robustness improvement that gives higher weight to those points in the images that contain more information. Finally, a fast parallel implementation using networked personal computers is presented. Results show that our method can take into account the large variability of the inner brain structures. A parallel implementation allowed us to execute the registration algorithm in 5 minutes and future improvements will open the possibility of registering massive quantities of images.

show abstract

Low Level Parallelization of Nonlinear Diffusion Filtering Algorithms for Cluster Computing Environments

Slogsnat

Fischer

Bruhn

et al. 2003

Euro-Par 2003 Parallel Processing

View full text Add to dashboard Cite

Designing 3-D Nonlinear Diffusion Filters for High Performance Cluster Computing

Cited by 7 publications

References 7 publications

High-Performance 3D Image Processing Architectures for Image-Guided Interventions

High-Performance 3D Image Processing Architectures for Image-Guided Interventions

Grid Enabled Non-rigid Registration with a Dense Transformation and a priori Information

Low Level Parallelization of Nonlinear Diffusion Filtering Algorithms for Cluster Computing Environments

Contact Info

Product

Resources

About