Medical Images Rician Based Denoising
A. Mart, Alba García Seco de Herrera, Emanuele Schiavi, Laboratorio de Anal, Juan Carlos · 2011
Resumen Medical image denoising and restoration are key steps for diagnosis and tracking of neurodegenerative processes such as Alzheimer and Parkinson diseases. We consider here a Rician-based denoising model suitable for structural MRI brain images. This model can be deduced using a Bayesian approach and it represents a modication of the model recently proposed in [1] where a Perona-Malik type operator is considered. Here we consider to regularize using the Total Variation operator. This provides a more robust mathematical and numerical treatment of the problem because of the well known properties of the Total Variation operator. The basic problem of this technique relies on the resolution of the following variational minimization problem: Let be a