Skip to main content
. 2022 Nov 18;36(2):547–561. doi: 10.1007/s10278-022-00718-4

Table 1.

Comparison of our FARNet with prior state-of-the-art methods on the cephalometric X-ray dataset with 19 annotated landmarks

Test1 data Test2 data
Methods Input size MRE 2mm 2.5mm 3mm 4mm MRE 2mm 2.5mm 3mm 4mm
Ibragimov et al. [20] - 1.84 71.70 77.40 81.90 88.00 - 62.74 70.47 76.53 85.11
Lindner et al. [25] - 1.67 74.95 80.28 84.56 89.68 1.92 66.11 72.00 77.63 87.42
Arik et al. [31] 800×640 - 75.37 80.91 84.32 88.25 - 67.68 74.16 79.11 84.63
Qian et al. [44] - - 82.50 86.20 89.30 92.60 - 72.40 76.15 79.65 85.90
Oh et al. [43] 800×640 1.18 86.20 91.20 94.40 97.70 1.44 75.89 83.36 89.26 95.73
Chen et al. [42] 800×640 1.17 86.67 92.67 95.54 98.53 1.48 75.05 82.84 88.53 95.05
Zhong et al. [38] 968×968 1.12 86.91 91.82 94.88 97.90 1.42 76.00 82.90 88.74 94.32
FARNet(Our) 800×640 1.12 88.03 92.73 95.96 98.48 1.42 77.00 84.42 89.47 95.21

The bold value in each column represents the best result