Geometry-driven diffusion smoothing of the MR-brain images using a novel variable conductance
Ivan Bajla, Igor Holländer · 2002
A novel variable conductance function for geometry-driven diffusion smoothing is developed. It is bared on the measure of the neighborhood anisotropism adopted from the adaptive linear convolution image filtering. The results of several smoothing methods, aimed at the improvement of 3D visualization of MRI tomograms of the brain, are demonstrated on a phantom study and by two measures of signal-to-noise ratio.