Frontiers in Neuroinformatics (Dec 2013)

Brain extraction using the watershed transform from markers

  • Richard eBeare,
  • Richard eBeare,
  • Jian eChen,
  • Jian eChen,
  • Christopher L Adamson,
  • Timothy eSilk,
  • Deanne K Thompson,
  • Deanne K Thompson,
  • Deanne K Thompson,
  • Joseph Y M Yang,
  • Joseph Y M Yang,
  • Vicki A Anderson,
  • Marc L Seal,
  • Amanda G Wood

DOI
https://doi.org/10.3389/fninf.2013.00032
Journal volume & issue
Vol. 7

Abstract

Read online

Isolation of the brain from other tissue types in magnetic resonance(MR) images is an important step in many types of neuro-imagingresearch using both humans and animal subjects. The importance ofbrain extraction is well appreciated - numerous approaches have beenpublished and the benefits of good extraction methods to subsequentprocessing are well known.We describe a tool - the marker based watershed scalper (MBWSS)- for isolating the brain in T1-weighted MR images built usingfiltering and segmentation components from the Insight Toolkit (ITK)framework. The key elements of MBWSS - the watershed transform frommarkers and aggressive filtering with large kernels - are techniquesthat have rarely been used in neuroimaging segmentation applications. MBWSSis able to reliably isolate the brain without expensive preprocessingsteps, such as registration to an atlas, and is therefore useful asthe first stage of processing pipelines. It is an informative exampleof the level of accuracy achievable without using priors in the formof atlases, shape models or libraries of examples.We validate the MBWSS using a publicly available dataset, a paediatriccohort, an adolescent cohort, intra-surgical scans and demonstrateflexibility of the approach by modifying the method to extract macaquebrains.

Keywords