Biomedical Image Processing and 3D Surface Reconstruction
Goal: To translate 2D medical image to 3D Surface Geometry
This project is part of my master's thesis research, medical image such as CT, MRI or color Doppler are being processed. Image segmentation is employed to determine the overall lumen geometry for each axial image. The lumen is isolated from each image before lumen edge detection, which is obtained by use of the discretized Laplace equation. Once the lumen edge is detected, radial basis functions (RBFs) are used to obtain a mathematical representation of the three-dimensional lumen geometry base on a cluster of edge pixels.

Automated Mesh Generation of a Bifurcation Geometry
Goal: To automate the mesh generation process for bifurcation geometry.
This project is the main part of my master's thesis research. Isotherms, which provide element-cutting planes that are nonintersecting and orthogonal to the isolated boundary are obtained by the solution of three conduction heat transfer problems on the bifurcation geometry using the RBFs. A high quality unstructured hexahedral mesh is then generated based on this information.