Skip to main content
. 2017 Nov 20;17:607–615. doi: 10.1016/j.nicl.2017.11.015

Table 2.

Analysis of the classifier performance for different sizes. Results for mean detection TPF, FPFDSCd and mean segmentation DSCs.

Method TPF FPF DSCd DSCs
Small lesions (3–10)
LR-NDFNB 11.76 30.0 0.08 0.08
LR-NDF 11.76 25.0 0.12 0.10
LR-DFNB 28.13 25.56 0.25 0.21
LR-DF 34.40 24.09 0.26 0.24
Medium lesions (11–50)
LR-NDFNB 40.84 9.16 0.45 0.29
LR-NDF 40.83 9.21 0.46 0.30
LR-DFNB 61.52 12.65 0.65 0.39
LR-DF 65.70 12.50 0.67 0.39
Large lesions (50+)
LR-NDFNB 77.80 11.76 0.81 0.49
LR-NDF 77.80 11.11 0.81 0.50
LR-DFNB 91.24 6.25 0.93 0.57
LR-DF 91.30 5.88 0.93 0.59