Real-time 3-D ultrasound scan conversion using a multicore processor

Authors:
Bo Zhuang;Vijay Shamdasani;Siddhartha Sikdar;Ravi Managuli;Yongmin Kim
Affiliations:
Ultrasonix Medical Corporation, Richmond, BC, Canada and Department of Bioengineering, University of Washington, Seattle, WA;Philips Medical Systems, Inc., Bothell and Department of Bioengineering, University of Washington, Seattle, WA;Department of Electrical and Computer Engineering, George Mason University, Fairfax, VA and Department of Bioengineering, University of Washington, Seattle, WA;Department of Bioengineering, University of Washington, Seattle, WA and Hitachi Medical Systems of America, Twinsburg, OH;Department of Bioengineering, University of Washington, Seattle, WA
Venue:
IEEE Transactions on Information Technology in Biomedicine - Special section on biomedical informatics
Year:
2009

Citing 7
Cited 0

Image interpolation and resampling

Handbook of medical imaging
Shear-image order ray casting volume rendering

I3D '03 Proceedings of the 2003 symposium on Interactive 3D graphics
Strategies for Mapping Algorithms to Mediaprocessors for High Performance

IEEE Micro
Chip Multithreading: Opportunities and Challenges

HPCA '05 Proceedings of the 11th International Symposium on High-Performance Computer Architecture
Industry Trends: Chip Makers Turn to Multicore Processors

Computer
Programmable ultrasound imaging using multimedia technologies: a next-generation ultrasound machine

IEEE Transactions on Information Technology in Biomedicine
A single mediaprocessor-based programmable ultrasound system

IEEE Transactions on Information Technology in Biomedicine

Quantified Score

Hi-index	0.00

Visualization

Abstract

Real-time 3-D ultrasound scan conversion (SC) in software has not been practical due to its high computation and I/O data handling requirements. In this paper, we describe software-based 3-D SC with high volume rates using a multicore processor, Cell. We have implemented both 3-D SC approaches: 1) the separable 3-D SC where two 2-D coordinate transformations in orthogonal planes are performed in sequence and 2) the direct 3-D SC where the coordinate transformation is directly handled in 3-D. One Cell processor can scan-convert a 192 × 192 × 192 16-bit volume at 87.8 volumes/s with the separable 3-D SC algorithm and 28 volumes/s with the direct 3-D SC algorithm.