Skip to main content
. 2021 Mar 18;443:96–105. doi: 10.1016/j.neucom.2021.03.034

Fig. 8.

Fig. 8

Visualization results of ResUNet. The images from left to right represent the original CXR images, masks predicted by trained ResUNet, ground-truth masks, and segmentation, for a (top) normal and a (bottom) TB cases in the test set.