Adaptive heal-doming can significantly improve the image quality in biomedical ultrasound by reducing the clutter due to interfering signals arriving from undesired directions. Adaptive beamforming is computationally expensive, and the objective of this thesis is to expose and explore tradeoffs between computational complexity and quality of adaptive beamforming. We consider the conventional linearly constrained minimum variance (LCMV) adaptive beamformer, applied to B-mode ultrasound imaging, and study an alternative based on the well-known generalized sidelobe canceller (GSC) whose adaptation relies on unconstrained gradient-driven optimization. To our knowledge, this is the first time a GSC-based gradient-driven approach has been applied and evaluated in the context of ultrasound beamforming. As another alternative to the conventional LCMV method, we also propose and evaluate a simple idea of updating the beamformer's weight vector at a reduced rate. Both approaches have lead to significant computational savings, but they also sacrifice beamforming optimality. Our simulations show that despite suboptimal beamforming. the ultrasound image quality remains acceptable.
| Identifer | oai:union.ndltd.org:uvic.ca/oai:dspace.library.uvic.ca:1828/3096 |
| Date | 15 November 2010 |
| Creators | Khezerloo, Solmaz |
| Contributors | Rakhmatov, Daler N. |
| Source Sets | University of Victoria |
| Language | English, English |
| Detected Language | English |
| Type | Thesis |
| Rights | Available to the World Wide Web |
Page generated in 0.0022 seconds