An MR image segmentation algorithm based on bias field correction

Bias field which existed in MR image had an unfavorable influence on analyzing medical image, especially for medical image segmentation. To solve it, this paper proposed an approach which is based on energy minimization to estimate bias field where exists in the MR image, finish brain tissue segmentation, meanwhile. This method fully utilize the properties of factual information and bias information in the MR image, and by means of effective and stabilized matrix iterative calculation we can obtain optimal value of bias field and every tissues. Furthermore, this method also reflects excellent, robustness and accuracy.