COMPARATIVE ACCURACY OF A NOVEL APPROACH TO AUTOMATIC TEMPORAL EVENT DETECTION DURING DROP VERTICAL JUMP
Jacob M. Thomas1, Jamie B. Hall1, Rebecca A. Bliss1, Trent M. Guess1
1University of Missouri, Columbia, Missouri
Motion capture technology can add meaningful insight to the drop vertical jump (DVJ); however, these systems are quite expensive. PURPOSE: Therefore, the purpose of this study was to compare a low-cost depth-sensing camera (DSC) with body tracking software to the gold standard optical retroreflective motion capture system (OMC) in detecting the timing of initial ground contact (IGC) and maximum knee flexion (MKF) using a novel approach to automatic IGC detection without a force plate. This approach was also validated with force plate IGC detection. METHODS: This pilot study included 19 individuals (24.16±2.39 yrs, 174.66±9.42 cm, 72.79±10.13 kg). Participants were fitted with retroreflective markers and completed 5 DVJ repetitions from a 31 cm box onto 2 force plates (right foot, left foot) sampling at 1000 Hz. Each trial was simultaneously recorded using a 12-camera OMC sampling at 100 Hz, and a DSC sampling at 30 Hz placed 2.7 m from the capture space, and 1.0 m above the ground in the frontal plane. A custom MATLAB (version 2019b) algorithm detected the moment left and right ankle joint centers crossed below 180 mm (relative to ground) in the vertical plane for both devices. Right and left times were averaged to determine IGC. Maximum right and left knee flexion values within 1 second following IGC were averaged to determine MKF for both devices. Force plate IGC was defined as the moment vertical ground reaction force exceeded 10% of participant body mass. Because the DSC and OMC could not be synchronized, time difference between MKF and IGC was analyzed as a measure of accuracy between the two. For this study, results of 1 trial were analyzed. ICC values measured agreement in time differences for all systems. Statistical analysis was performed in RStudio (version 4.1.1). RESULTS: ICC values indicated perfect agreement between kinematic calculation of IGC using the OMC and force plate calculation of IGC (ICC(3,k) = .99, p<.001). Average difference between MKF and IGC for the OMC was 0.29±0.12 seconds. For the DSC, average difference was 0.28±0.12 seconds. ICC values were very high between the DSC and OMC (ICC(3,k) = 0.99, p<.001) when comparing time between MKF and IGC. CONCLUSION: As this was a pilot study, findings are limited in their generalizability due to a reduced sample size. However, this work demonstrates that a single DSC can detect IGC and MKF events during DVJ, which could be useful for event detection in clinics where force plates are not available. Future studies should evaluate the usefulness of this technique using a greater and more diverse sample population, in order to further assess its accuracy.
ACKNOWLEDGEMENTS: This study was funded in part by the University of Missouri Coulter Biomedical Accelerator.
Thomas, JM; Hall, JB; Bliss, RA; and Guess, TM
"COMPARATIVE ACCURACY OF A NOVEL APPROACH TO AUTOMATIC TEMPORAL EVENT DETECTION DURING DROP VERTICAL JUMP,"
International Journal of Exercise Science: Conference Proceedings: Vol. 11:
9, Article 43.
Available at: https://digitalcommons.wku.edu/ijesab/vol11/iss9/43