Ma, D; Cardoso, M; Modat, M; Powell, N; Holmes, H; Lythgoe, M; Ourselin, S; (2012) Multi Atlas Segmentation applied to in vivo mouse brain MRI. In: Landman, BA and Warfield, SK, (eds.) MICCAI 2012 Workshop on Multi-Atlas Labeling. (pp. 134 - 143). MICCAI 2012 Workshop on Multi-Atlas Labeling: Nice.

PDF Ma. et al - 2012 - MICCAI 2012 Multi Atlas Workshop.pdf Available under License : See the attached licence file. Download (4MB)

Abstract

Multi-atlas segmentation propagation has evolved quickly in recent years, becoming a state-of-the-art method for automatic struc- tural parcellation for brain MRI. However, few studies have applied these methods to preclinical research. In this study, we present a fully auto- matic multi-atlas segmentation pipeline for mouse brain MRI tissue par- cellation. The pipeline adopts the Multi-STEPS multi-atlas segmentation algorithm, which utilises a locally normalised cross correlation (LNCC) similarity metric for atlas selection and an extended STAPLE frame- work for multi-label fusion. The segmentation accuracy of the pipeline was evaluated using an in vivo mouse brain atlas with pre-segmented manual labels as gold standard, and optimised parameters were obtained. Results show a mean Dice similarity coefficient of 0.839 over all the struc- tures and for all the samples in the database, significantly higher than in a single atlas propagation strategy, and also generally higher than STAPLE strategy, although the improvement is not significant.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as XMLJSONCSV

Archive Staff Only

View Item