A Cross-Center Smoothness Prior for Variational Bayesian Brain Tissue Segmentation

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

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.

Original languageEnglish
Title of host publicationInformation Processing in Medical Imaging : 26th International Conference, IPMI 2019, Proceedings
EditorsSiqi Bao, Albert C.S. Chung, James C. Gee, Paul A. Yushkevich
Publication date1 Jan 2019
ISBN (Print)9783030203504
Publication statusPublished - 1 Jan 2019
Event26th International Conference on Information Processing in Medical Imaging, IPMI 2019 - Hong Kong, China
Duration: 2 Jun 20197 Jun 2019


Conference26th International Conference on Information Processing in Medical Imaging, IPMI 2019
ByHong Kong
SeriesLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11492 LNCS

    Research areas

  • Bayesian transfer learning, Image segmentation, Variational inference


ID: 226392835