Bayesian VoxDRN: a probabilistic deep voxelwise dilated residual network for whole heart segmentation from 3D MR images

Shi, Zenglin, Zeng, Guodong, Zhang, Le, Zhuang, Xiahai, Li, Lei, Yang, Guang and Zheng, Guoyan (2018) Bayesian VoxDRN: a probabilistic deep voxelwise dilated residual network for whole heart segmentation from 3D MR images. Frangi, Alejandro F., Fichtinger, Gabor, Schnabel, Julia A., Alberola-López, Carlos and Davatzikos, Christos (eds.) In Medical Image Computing and Computer Assisted Intervention – MICCAI 2018 - 21st International Conference, 2018, Proceedings. vol. 11073 LNCS, Springer. pp. 569-577 . (doi:10.1007/978-3-030-00937-3_65).

Abstract

In this paper, we propose a probabilistic deep voxelwise dilated residual network, referred as Bayesian VoxDRN, to segment the whole heart from 3D MR images. Bayesian VoxDRN can predict voxelwise class labels with a measure of model uncertainty, which is achieved by a dropout-based Monte Carlo sampling during testing to generate a posterior distribution of the voxel class labels. Our method has three compelling advantages. First, the dropout mechanism encourages the model to learn a distribution of weights with better data-explanation ability and prevents over-fitting. Second, focal loss and Dice loss are well encapsulated into a complementary learning objective to segment both hard and easy classes. Third, an iterative switch training strategy is introduced to alternatively optimize a binary segmentation task and a multi-class segmentation task for a further accuracy improvement. Experiments on the MICCAI 2017 multi-modality whole heart segmentation challenge data corroborate the effectiveness of the proposed method.

This record has no associated files available for download.

Contributors

Author: Lei Li

Download statistics