Electrical and Computer Engineering
University of Victoria
Green Computing Innovations for Biomedical Imaging Applications
- Low-complexity image formation techniques for biomedical ultrasound applications, aiming to reduce the energy cost of multidimensional data acquisition, processing, and analysis;
- Configurable hardware/software architectures for supporting computation-intensive imaging tasks, to enable high-performance, low-power, and flexible application execution;
- Predictive analytical models of nonlinear dynamics of the system energy supply and demand, to facilitate battery-aware system-level optimizations (e.g., hardware/software partitioning, data precision, operation scheduling);
- Model-driven configuration algorithms that automatically tune hardware/software component organization and programming, to maximize system service quality subject to battery-imposed energy constraints during targeted application execution.
PUBLICATIONS