Professor Wan's research is in the area of scientific computing with interests in developing efficient numerical techniques for solving large sparse linear systems arising from partial differential equation modeling. He is also interested in scalable algorithms on multi-core processing architectures. Recently, Professor Wan has been applying these methods to physically-based animation, medical simulation of the brain and soft tissues, medical image processing, and computational finance.
Realistic image synthesis and simulations have been used for a wide variety of applications, e.g. CAD, architectural design, remote sensing, movies, fine arts, etc. As the demand for higher quality of realism increases, there has been recent interest in applying physical and models to computer simulations, especially in the visualization of physical phenomena. Example projects are simulations of aurorae, tornadoes, plant leaf growth, and fracturing of rigid bodies during explosions.
Professor Wan has also applied computer simulations to study the biomechanics of the brain in different medical conditions. In hydrocephalus, the excess accumulation of cerebrospinal fluid (CSF) inside the ventricle can lead to fatal damage of the brain. Professor Wan has developed a finite element viscoelastic model to study the mechanical behaviour of the brain as well as to develop a model to predict the outcome of a stunt surgery. In the study of traumatic brain injury, Professor Wan developed a numerical model to simulate the role of CSF in brain injury. He used the numerical results to explain the contre-coup phenomenon and verify the validity of other theories.
The biomedical research has recently been extended from modeling and simulation to medical image analysis applicable to image guided-procedures and pathological findings. One current project is on real-time 2D-3D image registration for cardiovascular image analysis, in collaboration with the group at Heart and Circulation Program, Sunnybrook Health Sciences Centre. Professor Wan is developing robust models and real-time tools utilizing the multi-core technology for more accurate and efficient image registration process. Another project is on tracking living cells from a sequence of images obtained from phase contrast microscopy to study the dynamics of biological systems and the fundamental processes such as cell metabolism, movement, differentiation and death. Professor Wan's group is currently developing robust image segmentation models suitable for low contrast images as well as tracking models for capturing cell motions. This project is in collaboration with researchers in Ontario Institute for Cancer Research and McGill University.
Mathematical modeling often leads to a set of complex (non-linear) equations to be solved accurately and efficiently. Consequently, new numerical algorithms are required to deal with the new and different applications. Professor Wan's research has been focused on developing robust and scalable methods for solving different equations. In particular, he designs new multigrid techniques to solve interface problems, convection dominated problems, and hyperbolic fluid flow problems. Recently, Professor Wan is interested in developing numerical techniques for solving equations arising from computational finance.
Degrees and awards
BS (Chinese U of Hong Kong), PhD (California, Los Angeles)
Tier II Canada Research Chair in Scientific Computing (2006)
Industrial and sabbatical experience
In 1995, Professor Wan was a Givens Fellows with the Mathematics and Computer Science Group, Argonne National Laboratory, where he developed parallel incomplete LU factorization preconditioning code for BlockSolve95, a parallel software system for solving large linear systems on high performance computers.
After graduating in 1998, Professor Wan was a Forsythe Fellow with the Computer Science Department, Stanford University. He was involved in the ASCI project, funded by DOE as one of 5 ASCI centers in US, for the simulation of turbine engines. After joining the University of Waterloo, Professor Wan has been invited to speak at the International Conference on Preconditioning Techniques for Large Sparse Matrix Problems in Industrial Applications, Minneapolis, 1999, and in the International Conference on Domain Decomposition Methods, Mexico, 2002.
Z. Yi and J.W.L. Wan. An Inviscid Model for Nonrigid Image Registration. Accepted in Inverse Problems and Imaging in 2010.
L. Bradbury and J.W.L. Wan. A Spectral K-Means Approach to Bright-field Cell Image Segmentation. Proceedings of Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2010.
L. Xu and J.W.L. Wan. Real-Time Intensity-Based Rigid 2D-3D Medical Image Registration Using RapidMind Multi-Core Development Platform. Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, 2008.
I.R. Wang, J.W.L. Wan, and P.A. Forsyth. Robust Numerical Valuation of European and American Options under the CGMY Process. Journal of Computational Finance, 10:31-69, 2007.
R. Bank, J.W.L. Wan, and Z. Qu. Kernel Preserving Multigrid Methods for Convection-Diffusion Equations. SIAM Journal on Matrix Analysis and Applications, 27:1150-1171, 2006.