•  
  •  
 

Abstract

International Journal of Exercise Science 17(1): 1629-1647, 2024. In weightlifting, quantitative kinematic analysis is essential for evaluating snatch performance. While marker-based (MB) approaches are commonly used, they are impractical for training or competitions. Markerless video-based (VB) systems utilizing deep learning-based pose estimation algorithms could address this issue. This study assessed the comparability and applicability of VB systems in obtaining snatch kinematics by comparing the outcomes to an MB reference system. 21 weightlifters (15 Male, 6 Female) performed 2–3 snatches at 65%, 75%, and 80% of their one-repetition maximum. Snatch kinematics were analyzed using an MB (Vicon Nexus) and VB (Contemplas along with Theia3D) system. Analysis of 131 trials revealed that corresponding lower limb joint center positions of the systems on average differed by 4.7 ± 1.2 cm, and upper limb joint centers by 5.7 ± 1.5 cm. VB and MB lower limb joint angles showed highest agreement in the frontal plane (root mean square difference (RMSD): 11.2 ± 5.9°), followed by the sagittal plane (RMSD: 13.6 ± 4.7°). Statistical Parametric Mapping analysis revealed significant differences throughout most of the movement for all degrees of freedom. Maximum extension angles and velocities during the second pull displayed significant differences (p < .05) for the lower limbs. Our data showed significant differences in estimated kinematics between both systems, indicating a lack of comparability. These differences are likely due to differing models and assumptions, rather than measurement accuracy. However, given the rapid advancements of neural network-based approaches, it holds promise to become a suitable alternative to MB systems in weightlifting analysis.

Additional Files
Ankle_dorsiflexion.png (180 kB)
Ankle plantarflexion/dorsiflexion angle (Mean ± SD) of both body sides across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems.
Ankle_inversion.png (145 kB)
Ankle inversion/eversion angle (Mean ± SD) of both body sides across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems.
Ankle_toe_out.png (172 kB)
Foot toe-out/toe-in angle (Mean ± SD) of both body sides across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems.
Hip_abduction.png (163 kB)
Hip abduction/adduction angle (Mean ± SD) of both body sides across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems.
Hip_external_rotation.png (176 kB)
Hip external/internal rotation angle (Mean ± SD) of both body sides across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems.
Knee_abduction.png (174 kB)
Knee abduction/adduction angle (Mean ± SD) of both body sides across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems.
Figure 1.png (804 kB)
Figure 1. Example shown the right knee of a trial of one athlete. Knee angle and angular extension velocity throughout the snatch computed with the MB and VB system. The vertical lines indicate the respective end or beginning of a phase. (1) First Pull; (2) Transition; (3) Second Pull; (4) Turnover; (5) Recovery.
Figure 2.png (17646 kB)
Figure 2. Pose estimation of the video-based motion capture system [A]. Plug-in-Gait model used with the marker-based motion capture system [B]. Lab setup with cameras [C].
Figure 3.png (188 kB)
Figure 3. Averaged right and left wrist position (Mean ± SD) across all trials and participants measured by the VB system (blue) and of the position of the averaged right and left end of the barbell measured by the MB system (green) are shown in the left column. The positions are relative to the position of the wrist (for the VB system) and the barbell (for the MB system) at the beginning of the snatch. The right column displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame.
Figure 4.png (179 kB)
Figure 4. Knee extension/flexion angles (Mean ± SD) across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems. The x-axis represents the time-normalized snatch movement for all rows.
Figure 5.png (1310 kB)
Figure 5. Lower extremity angles (Mean ± SD) across all trials for the snatch. The right y-axis shows the root mean square difference (RMSD) between the two systems calculated for each frame (yellow dotted line). Below each subplot of the joint angles and RMSD, the results of the SPM with t-values representing the y-axis are displayed. The red dashed lines indicate the threshold for significant differences determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The x-axis represents the time-normalized snatch movement for all rows.
Figure 6.png (949 kB)
Figure 6. Bland-Altman plot, illustrating the differences between the individual estimates of the maximum attained extension angles (first row) and the maximum attained extension velocity (second row) of both systems against the average of the measurements during the second pull. The solid line (black) represents the bias (mean difference), and dashed lines (red) represent the upper and lower limits of agreement. Each subplot includes the results of a two-sample T-test comparing the measurements of the two systems.
Figure 7.png (183 kB)
Figure 7. Hip extension/flexion angle (Mean ± SD) across all trials divided into the time-normalized phases computed with the VB (blue) and MB (green) motion capture system are shown in the first row. The second row displays the results of the SPM with t-values representing the y-axis. The red dashed lines indicate the threshold for significant differences, as determined by SPM pair-wise comparisons. T-values laying within the dashed lines suggest no significant differences between the two systems at the corresponding variable and frame. The third row represents the absolute difference (Mean ± SD) between the estimates of the two systems. The x-axis represents the time-normalized snatch movement for all rows.
Download

Share

COinS