A Generative Probabilistic Model and Discriminative Extensions for Brain Lesion Segmentation - With Application to Tumor and Stroke

Research output: Contribution to journalJournal article – Annual report year: 2016Researchpeer-review

Documents

DOI

  • Author: Menze, Bjoern H.

    Massachusetts Institute of Technology, United States

  • Author: Van Leemput, Koen

    Image Analysis & Computer Graphics, Department of Applied Mathematics and Computer Science , Technical University of Denmark, Denmark

  • Author: Lashkari, Danial

    Massachusetts Institute of Technology, United States

  • Author: Riklin-Raviv, Tammy

    Ben-Gurion University of the Negev, Israel

  • Author: Geremia, Ezequiel

    INRIA Sophia Antipolis, France

  • Author: Alberts, Esther

    Technische Universität München, Germany

  • Author: Gruber, Philipp

    University Hospital Zurich, Switzerland

  • Author: Wegener, Susanne

    University Hospital Zurich, Switzerland

  • Author: Weber, Marc-André

    University Hospital Heidelberg, Germany

  • Author: Székely, Gabor

    Swiss Federal Institute of Technology Zurich, Switzerland

  • Author: Ayache, Nicholas

    INRIA Sophia Antipolis, France

  • Author: Golland, Polina

    Massachusetts Institute of Technology, United States

View graph of relations

We introduce a generative probabilistic model for segmentation of brain lesions in multi-dimensional images that generalizes the EM segmenter, a common approach for modelling brain images using Gaussian mixtures and a probabilistic tissue atlas that employs expectation-maximization (EM), to estimate the label map for a new image. Our model augments the probabilistic atlas of the healthy tissues with a latent atlas of the lesion. We derive an estimation algorithm with closed-form EM update equations. The method extracts a latent atlas prior distribution and the lesion posterior distributions jointly from the image data. It delineates lesion areas individually in each channel, allowing for differences in lesion appearance across modalities, an important feature of many brain tumor imaging sequences. We also propose discriminative model extensions to map the output of the generative model to arbitrary labels with semantic and biological meaning, such as "tumor core" or "fluid-filled structure", but without a one-to-one correspondence to the hypo- or hyper-intense lesion areas identified by the generative model. We test the approach in two image sets: the publicly available BRATS set of glioma patient scans, and multimodal brain images of patients with acute and subacute ischemic stroke. We find the generative model that has been designed for tumor lesions to generalize well to stroke images, and the extended discriminative-discriminative model to be one of the top ranking methods in the BRATS evaluation.
Original languageEnglish
JournalI E E E Transactions on Medical Imaging
Volume35
Issue number4
Pages (from-to)933-946
ISSN0278-0062
DOIs
Publication statusPublished - 2016

Bibliographical note

This work is licensed under a Creative Commons Attribution 3.0 License.

CitationsWeb of Science® Times Cited: No match on DOI

    Research areas

  • Medical diagnostic imaging, Anatomical structure, Tumors, Image segmentation, Object segmentation, Bayes methods

Download statistics

No data available

ID: 124025714