Abstract
Targeting chemotherapy to cancer cells is crucial and remains to be a key objective in cancer treatments. Photodynamic Therapy (PDT) is an attempt to solve this problem by using light-sensitive drugs, called photosensitizers, to achieve selectivity. A critical step in PDT is the implantation of optical fibers into the tumor to irradiate the target volume, thereby activating these photosensitizers in the region of interest. However, the precise irradiation of the tumor requires thorough treatment planning to determine the optimal positions and emission profiles of the fibers, among other factors, in order to achieve a highly conformal therapy. One of the most flexible and accurate methods to compute the light dose distribution is the Monte Carlo method, which is unfortunately computationally intensive. To explore this possibility within the context of PDT treatment planning, the current project aims to accelerate a Monte Carlo simulation based on the Monte Carlo for Multi-Layered media (MCML) software on GPUs.
The Compute Unified Device Architecture (CUDA) has become a de-facto standard for programming NVIDIA GPUs. Although it is a simple extension to C, CUDA still places on the programmer the burden of packaging GPU code in separate functions, of explicitly managing data transfer between the host memory and various components of the GPU memory, and of manually optimizing the utilization of the GPU memory. hiCUDA is a high-level directive-based language for CUDA programming. It allows programmers to perform the above-mentioned tasks, often tedious and error-prone, in a simpler manner, and directly to the sequential code. At the same time, experiments using our prototype hiCUDA compiler show that hiCUDA does not sacrifice performance for ease-of-use, because the compiler-generated CUDA code can perform as well as the hand-written versions. The current project is a perfect opportunity to evaluate the effectiveness of hiCUDA.
Biography
William Lo received his B.A.Sc. degree in Computer Engineering from the University of Toronto in 2007. Currently, he is pursuing his M.Sc. degree in the Department of Medical Biophysics, under the guidance of Prof. Jonathan Rose and Prof. Lothar Lilge from the Ontario Cancer Institute at Princess Margaret Hospital.
David Han is a M.A.Sc. candidate in Computer Engineering at the University of Toronto, working under the supervision of Prof. Tarek Abdelrahman. He received his B.A.Sc. degree in Computer Engineering from the University of Toronto in 2007.