Abstract
Automated segmenting and labeling of individual brain anatomical regions, in MRI are challenging, due to the issue of individual structural variability. Although atlas-based segmentation has shown its potential for both tissue and structure segmentation, due to the inherent natural variability as well as disease-related changes in MR appearance, a single atlas image is often inappropriate to represent the full population of datasets processed in a given neuroimaging study. As an alternative for the case of single atlas segmentation, the use of multiple atlases alongside label fusion techniques has been introduced using a set of individual “atlases” that encompasses the expected variability in the studied population. In our study, we proposed a multi-atlas segmentation scheme with a novel graph-based atlas selection technique. We first paired and co-registered all atlases and the subject MR scans. A directed graph with edge weights based on intensity and shape similarity between all MR scans is then computed. The set of neighboring templates is selected via clustering of the graph. Finally, weighted majority voting is employed to create the final segmentation over the selected atlases. This multi-atlas segmentation scheme is used to extend a single-atlas-based segmentation toolkit entitled AutoSeg, which is an open-source, extensible C++ based software pipeline employing BatchMake for its pipeline scripting, developed at the Neuro Image Research and Analysis Laboratories of the University of North Carolina at Chapel Hill. AutoSeg performs N4 intensity inhomogeneity correction, rigid registration to a common template space, automated brain tissue classification based skull-stripping, and the multi-atlas segmentation. The multi-atlas-based AutoSeg has been evaluated on subcortical structure segmentation with a testing dataset of 20 adult brain MRI scans and 15 atlas MRI scans. The AutoSeg achieved mean Dice coefficients of 81.73% for the subcortical structures.
Highlights
Accurate segmentation of brain structures from magnetic resonance imaging (MRI) (Khan et al, 2011), functional MRI (Maldjian et al, 2003) and positron emission tomography (PET) (Tohka et al, 2007) is essential for quantitative studies of the brain, such as disease progression and aging
EXPERIMENTAL RESULTS We have applied the AutoSeg segmentation software pipeline to the brain MRI data set with 20 testing scans and 15 atlases
In our multi-atlas-based method, we achieved a mean Dice similarity coefficient (DSC) of 81.50% for caudate and putamen and 81.18% for amygdala, caudate, hippocampus, lateral ventricle, pallidus, and putamen. The experiments of these studies were conducted on different datasets, the relatively large improvement in results with our method indicates the advantage of using multiple atlases [see other multi-atlas segmentation papers such as (Asman and Landman, 2013)]
Summary
Accurate segmentation of brain structures from magnetic resonance imaging (MRI) (Khan et al, 2011), functional MRI (fMRI) (Maldjian et al, 2003) and positron emission tomography (PET) (Tohka et al, 2007) is essential for quantitative studies of the brain, such as disease progression and aging. Manual brain anatomical labeling (identification of anatomical brain structures and assignment of a unique label to each structure) is considered the most accurate means of giving the most accurate results closest to the true segmentation of brain structures. As the size and availability of large MRI databases increase, manual segmentation of brain structures is not realistic means of segmenting the brain because of the significant time-cost of human raters and unpredictable intra- and inter-rater variability. Automated segmentation methods are highly desirable when the size of MRI databases is considerably large (e.g., >50 cases). Automated anatomical brain region segmentation (labeling) of subcortical regions in MRI data is challenging, since the contrast between tissues is often low for a variety of brain structures (“Subcortical regions” is included since folding/shape variation may be a bigger determining factor than contrast for labeling cortical structures.). The commonly present shape and intensity variations in a number of diseases further complicate robust brain segmentation
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
