Segmenting MR Images Through Texture Extraction and Multiplicative Components Optimization

The segmentation of MRI is a challenging task due to artifacts introduced by the acquisition process, like bias field and noise. In this paper, using a cartoon-texture decomposition of the image, we present a strategy that segments the cartoon processed by simultaneous bias correction and denoising. Preliminary numerical tests show that our method is effective in segmenting MRI data corrupted by noise.