Skip to main content
. 2022 Nov 30;24(12):1755. doi: 10.3390/e24121755

Figure 7.

Figure 7

Visual comparisons between the proposed modules on the test set. Each row presents a nodule segmentation case and each column is the segmentation result of adding different components: (a) backbone. (b) backbone + MF. (c) backbone + MF +MD. (d) backbone + MF + MD + CM without loss strategy for weighted pixels located at edges. (e) backbone + MF + MD + CM with loss strategy for weighted pixels located at edges. GT: Ground-Truth.