High-resolution 3D visualization of human hearts with emphases on the cardiac conduction system components-a new platform for medical education, mix/virtual reality, computational simulation.

Introduction: High-resolution digitized cardiac anatomical data sets are in huge demand in clinical, basic research and computational settings. They can be leveraged to evaluate intricate anatomical and structural changes in disease pathology, such as myocardial infarction (MI), which is one of the most common causes of heart failure and death. Advancements in high-resolution imaging and anatomical techniques in this field and our laboratory have led to vast improvements in understanding cardiovascular anatomy, especially the cardiac conduction system (CCS) responsible for the electricity of the heart, in healthy/aged/obese post-mortem human hearts. However, the digitized anatomy of the electrical system of the heart within MI hearts remains unexplored.

Methods: Five post-mortem non-MI and MI human hearts were obtained by the Visible Heart Laboratories via LifeSource, Minneapolis, MN, United States (with appropriate ethics and consent): specimens were then transported to Manchester University with an material transfer agreement in place and stored under the HTA 2004, UK. After performing contrast-enhanced micro-CT, a visualization tool (namely Amira) was used for 3D high-resolution anatomical visualizations and reconstruction. Various cardiovascular structures were segmented based on the attenuation difference of micro-CT scans and tissue traceability. The relationship between the CCS and surrounding tissues in MI and non-MI human hearts was obtained. 3D anatomical models were further explored for their use in computational simulations, 3D printing and mix/virtual reality visualization.

Results: 3D segmented cardiovascular structures in the MI hearts elicited diverse macro-/micro- anatomical changes. The key findings are thickened valve leaflets, formation of new coronary arteries, increased or reduced thicknesses of pectinate and papillary muscles and Purkinje fibers, thinner left bundle branches, sinoatrial nodal atrophy, atrioventricular conduction axis fragmentation, and increased epicardial fat in some hearts. The propagation of the excitation impulses can be simulated, and 3D printing can be utilized from the reconstructed and segmented structures.

Discussion: High-resolution digitized cardiac anatomical datasets offer exciting new tools for medical education, clinical applications, and computational simulation.