An End-to-end Approach to Semantic Segmentation with 3D CNN and Posterior-CRF in Medical Images

TY - CONF

T1 - An End-to-end Approach to Semantic Segmentation with 3D CNN and Posterior-CRF in Medical Images

AU - Chen, Shuai

AU - Bruijne, Marleen de

PY - 2018/11/8

Y1 - 2018/11/8

N2 - Fully-connected Conditional Random Field (CRF) is often used as post-processing to refine voxel classification results by encouraging spatial coherence. In this paper, we propose a new end-to-end training method called Posterior-CRF. In contrast with previous approaches which use the original image intensity in the CRF, our approach applies 3D, fully connected CRF to the posterior probabilities from a CNN and optimizes both CNN and CRF together. The experiments on white matter hyperintensities segmentation demonstrate that our method outperforms CNN, post-processing CRF and different end-to-end training CRF approaches.

AB - Fully-connected Conditional Random Field (CRF) is often used as post-processing to refine voxel classification results by encouraging spatial coherence. In this paper, we propose a new end-to-end training method called Posterior-CRF. In contrast with previous approaches which use the original image intensity in the CRF, our approach applies 3D, fully connected CRF to the posterior probabilities from a CNN and optimizes both CNN and CRF together. The experiments on white matter hyperintensities segmentation demonstrate that our method outperforms CNN, post-processing CRF and different end-to-end training CRF approaches.

KW - cs.CV

UR - https://sites.google.com/view/med-nips-2018/abstracts

M3 - Paper

ER -