Abstract
Multi-atlas segmentation propagation has evolved quickly in recent years, becoming a state-of-the-art methodology for automatic parcellation of structural images. However, few studies have applied these methods to preclinical research. In this study, we present a fully automatic framework for mouse brain MRI structural parcellation using multi-atlas segmentation propagation. The framework adopts the similarity and truth estimation for propagated segmentations (STEPS) algorithm, which utilises a locally normalised cross correlation similarity metric for atlas selection and an extended simultaneous truth and performance level estimation (STAPLE) framework for multi-label fusion. The segmentation accuracy of the multi-atlas framework was evaluated using publicly available mouse brain atlas databases with pre-segmented manually labelled anatomical structures as the gold standard, and optimised parameters were obtained for the STEPS algorithm in the label fusion to achieve the best segmentation accuracy. We showed that our multi-atlas framework resulted in significantly higher segmentation accuracy compared to single-atlas based segmentation, as well as to the original STAPLE framework.
Highlights
Modified mice are widely used in the preclinical studies of human brain diseases such as Alzheimer’s disease, as they share more than 85% of their genes with humans [1]
We compared the segmentation accuracy obtained from our pipeline with the result obtained from a single-atlas segmentation propagation method, and with the simultaneous truth and performance level estimation (STAPLE) algorithm
We adopted the in vivo mouse brain MRI data from another atlas database, the National University of Singapore (NUS) mouse atlas, as test images, and validated the ability of our multi-atlas framework to parcellate unlabelled new data from different site
Summary
Modified mice are widely used in the preclinical studies of human brain diseases such as Alzheimer’s disease, as they share more than 85% of their genes with humans [1]. Worldwide efforts to understand the role of genes in brain morphology demand efficient data acquisition and analysis framework to quantify the consequences of gene function in development and pathology. High resolution MRI techniques (voxel size ,100 mm) are becoming an increasingly popular tool to study morphometric changes in transgenic mice. Large scale MRI phenotyping studies demand high-throughput acquisition and analysis of high-resolution 3D data. Automatic, accurate quantitative methods for MR image analysis are essential for effective phenotyping. Structural parcellation is a quantitative analysis method, which enables the morphometric characterisation of brain structures, such as shape and volume. Different automatic algorithms have been developed to overcome these limitations and meet the challenge of objective and accurate high throughput analysis [5,6]
Sign-in/Register to view highlights & summary 
Published Version (
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