As an early-start PhD student, I worked on expanding image segmentation methods for use within medical image analysis. In particular, I concentrated on level set algorithms and the influence of shape priors on segmentation performance. All algorithms were developed in three dimensions using C++/ITK. Experiments were performed on cone-beam computed tomography (CBCT) image sequences acquired for prostate radiotherapy at the Yale-New Haven Hospital.

With partial funding from NSERC, I worked under the supervision of Professor Konstantinos N. Plataniotis over the period of sixteen weeks on a biomedical signal processing project.

I investigated the effect of nonlinearities in state estimation problems, particularly the importance of initial conditions in extended Kalman filtering. I developed and implemented a new Kalman-variant filter, the multiple extended Kalman filter (M-EKF), for instantaneous physiological tremor frequency tracking from neural microelectrode recordings (MER).

This project resulted in the submission and acceptance of a four page paper, "A New Stochastic Estimator for Tremor Frequency Tracking", to the refereed IEEE International Conference on Acoustics, Speech, and Signal Processing, April 2009. A preprint can be found on my publications page.

With partial funding from NSERC, I worked under the supervision of Professor Konstantinos N. Plataniotis over the period of sixteen weeks on a video camera based location estimation project.

I developed a complete visual tracking system using three off-the-shelf network cameras, capable of tracking and estimating the location of multiple users in indoor environments. I researched and implemented techniques including foreground segmentation, camera calibration, occlusion resolution and Kalman filtering.

I also presented my research to faculty and students at the Undergraduate Engineering Research Day at the University of Toronto. A Quicktime movie of my presentation (with embedded videos) can be found on my talks page.