Table 4.
Studied vessel | Segmentation approach | End-diastolic frames corresponding to ECG estimations | End-diastolic frames not corresponding to ECG estimations | Missed cardiac cycles | P |
---|---|---|---|---|---|
All studied vessels | 1st-analyst 1st estimation | 2924 (90.4%) | 68 (2.1%) | 243 (7.5%) | |
1st-analyst 2nd estimation | 2940 (90.9%) | 66 (2.0%) | 227 (7.0%) | ||
2nd-analyst | 3020 (93.2%) | 72 (2.2%) | 147 (4.5%) | < 0.001 | |
CIB-methodology | 3036 (90.1%) | 202 (6.0%) | 131 (3.9%) | ||
DL-methodology | 2993 (93.4%) | 39 (1.2%) | 174 (5.4%) | ||
Left anterior descending artery | 1st-analyst 1st estimation | 1121 (93.5%) | 17 (1.4%) | 61 (5.1%) | |
1st-analyst 2nd estimation | 1182 (100.0%) | 0 (0.0%) | 0 (0.0%) | ||
2nd-analyst | 1121 (93.3%) | 19 (1.6%) | 61 (5.1%) | < 0.001 | |
CIB-methodology | 1111 (92.2%) | 23 (1.9%) | 71 (5.9%) | ||
DL-methodology | 1093 (91.6%) | 11 (0.9%) | 89 (7.5%) | ||
Left circumflex artery | 1st-analyst 1st estimation | 887 (86.7%) | 12 (1.2%) | 124 (12.1%) | |
1st-analyst 2nd estimation | 949 (89.8%) | 46 (4.4%) | 62 (5.9%) | ||
2nd-analyst | 965 (93.2%) | 24 (2.3%) | 46 (4.4%) | < 0.001 | |
CIB-methodology | 973 (90.4%) | 65 (6.0%) | 38 (3.5%) | ||
DL-methodology | 958 (94.0%) | 8 (0.8%) | 53 (5.2%) | ||
Right coronary artery | 1st-analyst 1st estimation | 916 (90.4%) | 39 (3.8%) | 58 (5.7%) | |
1st-analyst 2nd estimation | 809 (81.4%) | 20 (2.0%) | 165 (16.6%) | ||
2nd-analyst | 934 (93.1%) | 29 (2.9%) | 40 (4.0%) | < 0.001 | |
CIB-methodology | 952 (87.5%) | 114 (10.5%) | 28 (2.0%) | ||
DL-methodology | 942 (94.8%) | 20 (2.0%) | 32 (3.2%) |
CIB conventional image-based, DL deep learning, ECG electrocardiogram