Abstract:<p>Medical image is a cornerstone of modern healthcare providing unique diagnostic. However, intensity nonuniformity is quite often found in medical images and unavoidably poses many challenges for precise image segmentation. In this paper, we present an incorporating global multiplicative decomposition and local statistical information level set technique for tissue segmentation and bias estimation that handles both high noise and intensity nonuniformity. An image can be factored into two multiplicative parts (namely, bias-free image and bias field), which are used to capture global intensity statistics. At the same time, our technique considers statistical information with not only the mean but also the variance. Pixel intensity generally fluctuates around its mean in a local region. The variance can adjust the disagreement of each pixel to its mean, and is explored to overcome the effects of intensity nonuniformity. Finally, the level set energy formulation is in the form of the global multiplicative decomposition and local statistical information. To verify the availability of our technique, we have thoroughly experimented on composite and real images. Our techniques are also applied to the lesion regions of brains. Experimental results display that our technique can segment different kinds of images with high noise and intensity nonuniformity more accurately, and outperforms other solutions in comparison.</p>

A Multiphase Chan-Vese Model Using Edge Information and Bias Field Correction for Image Segmentation

Image Segmentation Based on the Hybrid Bias Field Correction

A New Active Contour Model Driven by Pre-Fitting Bias Field Estimation and Clustering Technique for Image Segmentation.

A level set method based on additive bias correction for image segmentation

Active Contour Model Based on Local Bias Field Estimation for Image Segmentation.

A Three-Stage Variational Image Segmentation Framework Incorporating Intensity Inhomogeneity Information

An enhanced multiphase Chan-Vese model for the remote sensing image segmentation

A convex neighbor-constrained active contour model for image segmentation

Active contour model based on adaptive segmentation and bias field estimation

Robust image segmentation and bias field correction model based on image structural prior constraint

A variational level set approach to segmentation and bias correction of images with intensity inhomogeneity.

A Belief Propagation Algorithm For Bias Field Estimation And Image Segmentation

Split Bregman Method for Minimization of Modified Vese-Chan Model for Fast Image Segmentation

Improved Level Set Model Based on Bias Information with Application to Color Image Segmentation and Correction

An improved level set method for fast image segmentation based on single parameter

A New Method for Simultaneous Moderate Bias Correction and Image Segmentation

A Variational Multiphase Level Set Approach to Simultaneous Segmentation and Bias Correction

Incorporating global multiplicative decomposition and local statistical information for brain tissue segmentation and bias field estimation

Efficient active contour model based on Vese-Chan model and split bregman method

Simultaneous Segmentation and Correction Model for Color Medical and Natural Images with Intensity Inhomogeneity

Improved Segmentation Model Combining Region and Edge Information for Inhomogeneous Images.