Table 8 Performance of the action quality assessment for seven different FMS movements.
From: Functional movement screen dataset collected with two Azure Kinect depth sensors
Movement | F1 score | Accuracy (%) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Folder1 | Folder2 | Folder3 | Folder4 | Folder5 | Folder1 | Folder2 | Folder3 | Folder4 | Folder5 | |
Deep squat | 0.728 | 0.761 | 0.770 | 0.736 | 0.534 | 80.8 | 84.7 | 84.6 | 85.2 | 66.0 |
Hurdle step | 0.629 | 0.610 | 0.730 | 0.681 | 0.622 | 66.7 | 75.0 | 77.6 | 83.3 | 72.0 |
In-line lunge | 0.420 | 0.537 | 0.677 | 0.836 | 0.435 | 79.2 | 87.3 | 83.1 | 92.6 | 79.2 |
Shoulder mobility | 0.379 | 0.367 | 0.365 | 0.345 | 0.196 | 52.0 | 46.7 | 37.5 | 38.1 | 38.5 |
Active straight raise | 0.725 | 0.739 | 0.657 | 0.757 | 0.664 | 83.3 | 77.1 | 70.2 | 80.7 | 72.2 |
Trunk stability push-up | 0.713 | 0.472 | 0.582 | 0.881 | 0.269 | 80.8 | 0.848 | 79.2 | 95.7 | 67.7 |
Rotary stability | 0.535 | 0.675 | 0.519 | 0.396 | 0.321 | 88.8 | 86.7 | 91.9 | 85.9 | 92.9 |
Overall | 0.685 | 0.699 | 0.702 | 0.693 | 0.608 | 78.2 | 81.8 | 80.7 | 83.4 | 77.1 |
