Electrical field approximations from computational simulations are used to develop heart treatment and diagnostic tools and the effectiveness of these simulations depends on the accuracy and detail of the geometric model on which they are based. Because of the difficulty in creating high quality geometric models, many simulations use generic heart geometries and associated fiber orientations (e.g., the University of Auckland heart) even though these models are generic. They do not take into account geometric variation among individual hearts, which can play a significant role in the formation and shape of the electrical fields. Improvements in imaging techniques and computational algorithms have allowed us to develop a pipeline to generate high fidelity, subject-specific models of the heart that remove many of the approximations used in previous models. In a study of myocardial ischemia, this pipeline was used to generate multi-material models of canine left and right ventricles, including the ischemic region as determined by injecting contrast agent into the occluded artery. High-resolution anatomical and diffusion tensor imaging (DTI) scans were obtained from a 7 tesla small animal MRI scanner. From these scans, isosurfaces bounding different tissue regions were created using Seg3D for both the normal and ischemic tissue. A volume tetrahedral mesh was produced from the Isosurfaces using the program Tetgen and fiber orientation was mapped to each element from the DTI data. The resulting mesh, with anisotropic tissue conductivities assigned based on previous studies will be the basis for a bidomain computation of cardiac potential during acute myocardial ischemia using the SCIRun problem solving software. Direct comparison with experimental measurements of cardiac potentials during controlled acute ischemia will provide direct validation of the simulations. The main achievement of this research is the novel ability to create within hours customized geometric models of the heart from MRI scans, which will allow the evaluation of the role of geometric variation and uncertainty on the accuracy of simulated cardiac potentials.