Table 4 The clinical and imaging characteristics used for the external validation of three regression models
External validation of model 1 (AB-SSNs vs. NB-SSNs) | ||
Characteristics | AB-SSNs (n = 35) | NB-SSNs (n = 30) |
Detected time | ||
Initial CT | 25 (71.43%) | 28 (93.33%) |
Follow-up CT | 10 (28.57%) | 2 (6.67%) |
Margin | ||
Well-defined | 5 (14.29%) | 18 (60.00%) |
Ill-defined | 30 (85.71%) | 12 (40.00%) |
External validation of model 2 (AB-SSNs vs. M-SSNs) | ||
Characteristics | AB-SSNs (n = 35) | M-SSNs (n = 86) |
Age (years) | ||
Average | 52 ± 16 | 53 ± 12 |
Range | 18–82 | 30–71 |
Respiratory symptoms | ||
Presence | 4 (11.43%) | 10 (11.63%) |
Absence | 31 (88.57%) | 76 (88.37%) |
Detected time | ||
Initial CT | 25 (71.43%) | 86 (100%) |
Follow-up CT | 10 (28.57%) | 0 (0%) |
Margin | ||
Well-defined | 5 (14.29%) | 76 (88.37%) |
Ill-defined | 30 (85.71%) | 10 (11.63%) |
Shape | ||
Regular | 32 (91.43%) | 68 (79.07%) |
Irregular | 3 (8.57%) | 18 (20.93%) |
Target sign | ||
Presence | 6 (17.14%) | 4 (4.65%) |
Absence | 29 (82.86%) | 82 (95.35%) |
CT value of GGO areas (HU) | − 722.00 ± 171.00 | − 606.22 ± 106.02 |
Thickened interlobular septa | ||
Presence | 2 (5.71%) | 2 (2.33%) |
Absence | 33 (94.29%) | 84 (97.67%) |
Bubble lucency | ||
Presence | 0 (0%) | 18 (20.93%) |
Absence | 35 (100%) | 68 (79.07%) |
External validation of model 3 (NB-SSNs vs. M-SSNs) | ||
Characteristics | NB-SSNs (n = 30) | M-SSNs (n = 86) |
Detected time | ||
Initial CT | 28 (93.33%) | 86 (100%) |
Follow-up CT | 2 (6.67%) | 0 (0%) |
Size (mm) | ||
Average | 8.00 ± 3.00 | 10.00 ± 3.00 |
Range | 5–14 | 5–19 |
Shape | ||
Regular | 19 (63.33%) | 68 (79.07%) |
Irregular | 11 (36.67%) | 18 (20.93%) |
Target sign | ||
Presence | 6 (20.00%) | 4 (4.65%) |
Absence | 24 (80.00%) | 82 (95.35%) |
CT value of GGO areas (HU) | − 661.50 ± 98.00 | − 606.22 ± 106.02 |