Visualisation of 2D cardiovascular velocity maps with the Hilbert transform

Visualisation of human blood flow has become an important tool to investigate hemodynamics. These visual techniques can significantly contribute to gain insight into human hemodynamics, through looking at flow patterns in arteries and cardiac chambers. Non invasive imaging modalities such as flow encoded Echo-Planar Imaging (flow EPI) has enabled us to create real time images with flow and anatomical information. However, flow EPI generates 2D and 3D velocity maps that are not easily visualised because of its poor Signal-to-Noise Ratio (SNR). We applied the Hilbert transform to our bidimensional flow maps to filter out those irrelevant details and keep the appropriate flow information. Images formed with the real and imaginary parts as well as the phase component are presented. We also produced velocity vector images with the real and imaginary parts obtained with the Hilbert-filtering process. Finally, it possible to appreciate that these vector images reveal patterns of blood flow and certain trends in a distinct manner. This method combined with flow EPI may develop into a robust technique to investigate the cardiovascular hemodynamics in humans from 2D velocity maps.