Skip to main content
. 2020 Jul 23;33(5):1144–1154. doi: 10.1007/s10278-020-00355-9

Fig. 8.

Fig. 8

Image generated by intermediate steps in the network. The first row is 3D images generated by marching cubes [30] algorithm (presented with size of 48×48×48). The second row is the same section in 3D images. a Input 3D GGN image. b The output image of attention stage 1. c The output image of attention stage 2. d The output image of attention stage 3