Skip to main content
. Author manuscript; available in PMC: 2021 Feb 1.
Published in final edited form as: Med Image Anal. 2019 Nov 23;60:101621. doi: 10.1016/j.media.2019.101621

Figure 2:

Figure 2:

Overview of the proposed context-guided multi-task fully convolutional network (FCN), including two sub-networks (i.e., FCN1, and FCN2). FCN1 estimates the displacement maps, while FCN2 performs joint bone segmentation and landmark digitization. Each sub-network contains a contracting patch and an expanding path.