Tensor Domain Averaging in Diffusion Imaging of Small Animals to Generate Reliable Tractography

Abstract

Testing on small animal models is roughly the only path to transfer science-based knowledge to human use. More avidly than other human organs, we study the brain through animal models due to the complexity of experimenting directly on human subjects, even at a cellular level where the skull makes tissue sampling harder than in any other organ. Thanks to recent technological advances in imaging, animals do not need to be sacrificed. Magnetic resonance, in particular, favors long-term analysis and monitoring since its methods do not perturb the organ functions nor compromise the metabolism of the animals. Neurons' integrity is now indirectly visible under specialized mechanisms that use water displacement to track static boundaries. Although these water diffusion methods have proven to be successful in detecting neuronal structure at the submillimeter scale, they yield noisy results when applied to the resolutions required by small animals or when facing low myeline contents as in neonates and young children. This manuscript presents a strategy to display neuronal trending representations that follow the corticospinal tract's pathway and neuronal integrity in small rodents. The strategy is the foundation to study human neurodegenerative diseases and neurodevelopment as well.