A cross-center smoothness prior for variational Bayesian brain tissue segmentation
This work addresses the challenge of generalizing tissue segmentation across medical centers with varying acquisition protocols, which is incremental as it builds on existing Bayesian models by incorporating a learned smoothness prior.
The paper tackles the problem of brain tissue segmentation in MR images across different medical centers by introducing a cross-center smoothness prior that is learned from segmentations at another center and integrated into an unsupervised Bayesian model, resulting in segmentations with consistent smoothness without requiring labeled training data at the target center.
Suppose one is faced with the challenge of tissue segmentation in MR images, without annotators at their center to provide labeled training data. One option is to go to another medical center for a trained classifier. Sadly, tissue classifiers do not generalize well across centers due to voxel intensity shifts caused by center-specific acquisition protocols. However, certain aspects of segmentations, such as spatial smoothness, remain relatively consistent and can be learned separately. Here we present a smoothness prior that is fit to segmentations produced at another medical center. This informative prior is presented to an unsupervised Bayesian model. The model clusters the voxel intensities, such that it produces segmentations that are similarly smooth to those of the other medical center. In addition, the unsupervised Bayesian model is extended to a semi-supervised variant, which needs no visual interpretation of clusters into tissues.