Development of whole-heart model-based cardiac shape and motion analysis using parallel computing multi-core CPUs and GPUs technology

According to Statistics Canada, cardiovascular disease is the leading cause of death in Canada,
accounting for 72,338 or 32% of all Canadian deaths in 2004. To further reduce the mortality and
morbidity rate of heart disease, significant improvements in the extraction of the clinical heart
parameters, such as the change in shape and motion patterns have to be made, which are believed to
be useful for the early and reliable diagnosis of regional ischemia and dysfunctional contraction. The
accurate calculation of these parameters depends on an effective generating of 3D heart model,
which however, is a very computation demanding process. This project is designed to develop an
efficient implementation of spherical harmonic modeling algorithm by incorporating the combined,
parallel computing power of multi-core central processing units and graphics processing units, leading
to massive reductions in processing time and furthermore, to provide highly informative and accurate
parameters for clinical use.

Faculty Supervisor:

Hung-Yu Lin

Student:

Partner:

MedVoxel

Discipline:

Life Sciences

Sector:

University:

University of Manitoba

Program:

Accelerate