Automatic Segmentation of Neonatal Images Using Convex Optimization and Coupled Level Set Method
Accurate segmentation of neonatal brain MR images remains challenging mainly due to poor spatial resolution, low tissue contrast, high intensity inhomogeneity. Most existing methods for neonatal brain segmentation are atlas-based and voxel-wise. Although parametric or geometric deformable models have been successfully applied to adult brain segmentation, to the best of our knowledge, they are not explored in neonatal images. In this paper, we propose a novel neonatal image segmentation method, combining local intensity information, atlas spatial prior and cortical thickness constraint, in a level set framework. Besides, we also provide a robust and reliable tissue surfaces initialization for our proposed level set method by using a convex optimization technique. Validation is performed on 10 neonatal brain images with promising results.
KeywordsWhite Matter Gray Matter Segmentation Result Automatic Segmentation Neonatal Brain
Unable to display preview. Download preview PDF.
- 9.Zeng, X., Staib, L., Schultz, R., Duncan, J.: Segmentation and measurement of the cortex from 3D MR images using coupled surfaces propagation. IEEE Trans. Med. Imag. 18(10), 100–111 (1999)Google Scholar
- 15.Li, C., Kao, C., Gore, J., Ding, Z.: Implicit active contours driven by local binary fitting energy. In: CVPR, pp. 1–7 (2007)Google Scholar
- 16.Paragios, N.: A variational approach for the segmentation of the left ventricle in mr cardiac images. In: VLSM 2001 (2001)Google Scholar
- 19.Goldstein, T., Bresson, X., Osher, S.: Geometric applications of the split bregman method: Segmentation and surface reconstruction. CAM Report 09-06, UCLA (2009)Google Scholar