Title :
MRI bias field estimation based on diffusion
Author :
Yanfei Chen ; Zhiping Dan ; Xi Chen ; Shuanghui Lei ; Xiaobo Liu
Author_Institution :
Sch. of Electr. & Inf. Eng., Wuhan Inst. of Technol., Wuhan, China
Abstract :
Aim at magnetic resonance imaging (MRI) inherent bias field, a new method to estimate image bias field based on diffusion was proposed, which was based on an intuitive assumption that the true bias field can be evolved and approximated by image surface. To constrain this evolution, an energy function was designed based on two diffusion constraints. One was that the bias-free image has concise representation with image gradients. The other constraint was that the model fit on the bias field which was more relaxed than previous methods. Experimental results show that this method can estimate adaptively and correct effectively magnetic resonance imaging bias field, and compared with other methods, this model was more simple, but this effect was better.
Keywords :
biodiffusion; biomedical MRI; image representation; medical image processing; MRI bias field estimation; bias-free image; diffusion constraints; energy function; image bias field based; image gradients; image surface; magnetic resonance imaging inherent bias field; model fit; true bias field; Automation; Biomedical imaging; Bismuth; Educational institutions; Estimation; Magnetic resonance imaging; TV; bias field; diffusion; energy function; magnetic resonance imaging;
Conference_Titel :
Intelligent Control and Automation (WCICA), 2014 11th World Congress on
DOI :
10.1109/WCICA.2014.7052941