Modeling of the human heart in 3D using DTI images

Modeling the heart requires detailed representation of both the myocardial structure and the conduction system. Traditional CT and MRI imaging techniques fall short in capturing the anisotropic characteristics essential for accurate modeling of the heart's electrical and mechanical behaviors. This study employs Diffusion Tensor Magnetic Resonance Imaging (DT-MRI) to extract and model the myocardial fiber structure in a macroscopic domain. The diffusion properties are analyzed using scalar indices such as Mean Diffusivity, Fractional Anisotropy, and Diffusion Volume, enabling differentiation between myocardial and non-myocardial tissues. The dataset, acquired from a healthy subject, is processed using thresholding techniques to segment the myocardium and reconstruct a 3D heart model. The resulting anatomical model is integrated into a full human torso to enable forward modeling of the Body Surface Potential Map (BSPM). This study lays the groundwork for simulating cardiac excitation and paves the way for future inclusion of the conduction system.