Using GPUs for Faster Motion Estimation in Medical Images
Author Information
Author(s): Thiyagalingam Jeyarajan, Goodman Daniel, Schnabel Julia A., Trefethen Anne, Grau Vicente
Primary Institution: University of Oxford
Hypothesis
Can GPU-specific implementations of motion estimation algorithms significantly improve performance in medical image analysis?
Conclusion
The study demonstrates that mapping an enhanced motion estimation algorithm to GPU architectures can lead to performance gains of up to 60 times, enabling near-real-time processing.
Supporting Evidence
- GPU implementations can lead to performance gains of up to 60 times.
- The study provides insights into the challenges and benefits of using GPUs for medical image analysis.
- Different GPU architectures were evaluated to assess their performance in motion estimation.
Takeaway
This study shows that using special computer chips called GPUs can make analyzing medical images much faster, helping doctors get results in real time.
Methodology
The study involved mapping a motion estimation algorithm to different GPU architectures and evaluating its performance using a database of 3D image sequences.
Limitations
The performance gains may vary based on the specific GPU architecture and the complexity of the algorithms used.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website