Table 2.
Cross-validation | Testing | ||||||||
---|---|---|---|---|---|---|---|---|---|
Per-video | Per-study (Nab_videos ≥ 2) | Per-video | Per-study (Nab_videos ≥ 2) | ||||||
Ground truth | Ground truth | Ground truth | Ground truth | ||||||
Abnormal | Normal | Abnormal | Normal | Abnormal | Normal | Abnormal | Normal | ||
DL | Abnormal | 131 | 23 | 33 | 5 | 126 | 23 | 32 | 3 |
Normal | 37 | 91 | 4 | 5 | 26 | 71 | 1 | 5 | |
Sens | 0.780 | Sens | 0.892 | Sens | 0.829 | Sens | 0.970 | ||
Spec | 0.798 | Spec | 0.500 | Spec | 0.755 | Spec | 0.625 | ||
Acc | 0.787 | Acc | 0.809 | Acc | 0.801 | Acc | 0.902 | ||
κ | 0.567 | κ | 0.407 | κ | 0.581 | κ | 0.657 |
Forty-seven CT studies with 40% < LVEF < 60% were in the cross-validation and 41 CT studies were in the testing. The light gray indicates per-video evaluation, dark gray indicates per-study evaluation.