Lesion Segmentation

Warfield, Simon K.; X., Tomas-Fernandez,

doi:10.1016/b978-0-12-397025-1.00302-x

Cited by 1 publication

(1 citation statement)

References 62 publications

(54 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Identifying and measuring the spatial and temporal disseminations of lesions are key components of diagnostic criteria for MS. Traditional approaches to segment MS lesions include the utilization of supervised classification [1] or unsupervised clustering [14,12,8] to separate lesions from the normal brain tissues, where the former are commonly modeled as either an additional class or the outliers to the latter.…”

Section: Introductionmentioning

confidence: 99%

Ensemble of Multi-sized FCNs to Improve White Matter Lesion Segmentation

Wang

Smith

Liu

2018

Machine Learning in Medical Imaging

View full text Add to dashboard Cite

In this paper, we develop a two-stage neural network solution for the challenging task of white-matter lesion segmentation. To cope with the vast variability in lesion sizes, we sample brain MR scans with patches at three different dimensions and feed them into separate fully convolutional neural networks (FCNs). In the second stage, we process large and small lesion separately, and use ensemble-nets to combine the segmentation results generated from the FCNs. A novel activation function is adopted in the ensemble-nets to improve the segmentation accuracy measured by Dice Similarity Coefficient. Experiments on MIC-CAI 2017 White Matter Hyperintensities (WMH) Segmentation Challenge data demonstrate that our two-stage-multi-sized FCN approach, as well as the new activation function, are effective in capturing white-matter lesions in MR images.

show abstract