MARKERLESS MOTION CAPTURE SYSTEM – INTER-DEVICE RELIABILITY

Dimitrije Cabarkapa1, Eric M. Mosier2, Justin X. Nicoll3, Patrick G. Moodie4, Andrew C. Fry1 1University of Kansas, Lawrence, KS; 2Northwest Missouri State University, Maryville, MO; 3California State University, Northridge, CA; 4Scientific Analytics, Lenexa, KS. Ground reaction forces from a markerless motion capture system (MCS) have been shown to be accurate (Fry et al. 2016), however the measurements must also be reliable. PURPOSE: The purpose of this project was to determine the inter-device reliability between two markerless MCSs by using two identical devices. METHODS: A three-dimensional markerless MCS (DARI, Lenexa, KS) was used to analyze kinetic and kinematic data. Identical MCSs (Device A and Device B), with identical camera placement, were set up and calibrated in the same area by the same person. Subjects (n=50, 33 male and 17 female; ±SD, age=28.2±7.4 yrs, hgt=67.2±5.1 in, wgt=173.7±42.6 lbs) performed a standardized 19-movement test designed to assess overall body functional motor capabilities. For each movement during the test, both devices were recording and collecting data simultaneously. System algorithms used 192 test variables to calculate five scores (arbitrary units); power, functional strength, dysfunction, composite (power+functional–dysfunction), and vulnerability. Our hypothesis was that both devices would produce similar scores. T-test, Intra-class Correlation Coefficients (ICC) and Bland-Altman plots with 95% confidence intervals (CI) were used to determine inter-device reliability. RESULTS: Results are presented in the Table 1. There were no significant differences (p≥0.49) for any scores between the two devices tested. All ICCs were highly significant (ICC=0.86–0.99). These results were further supported by Bland-Altman plots, where all mean difference values were well within 95% CI. CONCLUSION: Based on these results, the 3-dimensional markerless MCS exhibited excellent inter-device reliability