Table 8 The average DSC scores per video of the test patients (ranked worst to best) between the algorithm’s predicted segmentations and each expert ground truth segmentation labels. Each patient (P) in the test set included 2 videos labelled as a and b
From: Automatic deep learning-based pleural effusion segmentation in lung ultrasound images
P | Expert 1 / Expert 2 | P | Algorithm / Expert 1 | P | Algorithm / Expert 2 |
|---|---|---|---|---|---|
4b | 0.312 +/- 0.10 | 4b | 0.564 +/- 0.02 | 4b | 0.569 +/- 0.02 |
1a | 0.449 +/- 0.06 | 2b | 0.604 +/- 0.03 | 1a | 0.573 +/- 0.04 |
2b | 0.550 +/- 0.05 | 4a | 0.630 +/- 0.03 | 1b | 0.574 +/- 0.04 |
1b | 0.563 +/- 0.08 | 1b | 0.632 +/- 0.06 | 4a | 0.601 +/- 0.03 |
5b | 0.593 +/- 0.05 | 1a | 0.649 +/- 0.06 | 2b | 0.684 +/- 0.03 |
5a | 0.665 +/- 0.10 | 5a | 0.719 +/- 0.06 | 5b | 0.776 +/- 0.02 |
4a | 0.671 +/- 0.06 | 5b | 0.721 +/- 0.05 | 5a | 0.820 +/- 0.01 |
3b | 0.743 +/- 0.04 | 2a | 0.830 +/- 0.02 | 3b | 0.830 +/- 0.02 |
2a | 0.762 +/- 0.05 | 3b | 0.860 +/- 0.03 | 3a | 0.853 +/- 0.01 |
3a | 0.772 +/- 0.02 | 3a | 0.898 +/- 0.02 | 2a | 0.864 +/- 0.02 |