Skip to main content

Table 4 The number of cases and evaluation matrices of segmentation results for clinical application in each category

From: An automated liver segmentation in liver iron concentration map using fuzzy c-means clustering combined with anatomical landmark data

Output categories FCM combined with anatomical landmark data FCM
Training cohort Testing cohort All data All data
Output I
 No. cases 359 (81.41%) 87 (77.68%) 446 (80.65%) 349 (63.11%)
 %DSC 91.30 ± 3.05 91.70 ± 2.55   
 %JC 84.13 ± 5.04 84.78 ± 4.30   
 H (mm) 15.75 ± 9.66 15.38 ± 9.25   
 %FP 8.83 ± 6.96 6.82 ± 5.33   
 %FN-Type I 8.47 ± 7.20 10.02 ± 6.98   
 %FN-Type II 0.00 0.00   
Output II
 No. cases 51 (11.56%) 12 (10.71%) 63 (11.39%) 150 (27.12%)
 %DSC 86.43 ± 6.49 85.44 ± 7.14   
 %JC 76.63 ± 9.45 75.18 ± 10.50   
 H (mm) 32.50 ± 14.77 28.76 ± 14.68   
 %FP 14.13 ± 12.94 18.00 ± 12.35   
 %FN-Type I 11.84 ± 7.49 9.01 ± 5.55   
 %FN-Type II 0.59 ± 0.81 1.09 ± 1.79   
Output III
 No. cases 31 (7.03%) 13 (11.61%) 44 (7.96%) 54 (9.76%)
 %DSC 54.26 ± 30.74 47.31 ± 23.93   
 %JC 42.87 ± 27.91 33.85 ± 20.02   
 H (mm) 72.39 ± 45.58 96.26 ± 46.01   
 %FP 53.12 ± 32.01 63.14 ± 25.31   
 %FN-Type I 6.42 ± 6.81 5.28 ± 12.39   
 %FN-Type II 0.65 ± 1.80 0.37 ± 1.32   
Total 441 112 553 553