Cardiac Fiber Tracking Using Adaptive Sampling Based on Diffusion Fractional Anisotropy
Abstract
As a non-invasive detection technology, magnetic resonance diffusion imaging (MRI) has aroused increasingly more attention in cardiac fiber tracking. How to improve efficiency and accuracy is still a challenging problem in this field. We presented a probabilistic method using adaptive sampling, which adjust sample size and sample divergence based on diffusion anisotropy to characterize the probability distribution of fiber path. We validate the method on real cardiac data set and get better results with stronger robustness.