The Concurrent Validity and Test-Retest Reliability of a Smartphone-Based Markerless System

Increasing the accessibility and portability of precise biomechanical data to sports scientists can assist in making data-driven decisions. The purpose of this study was: (1) Assess the concurrent and convergence validity of discrete and continuous waveform biomechanics of a smartphone-based markerless system. (2) Assess test-retest reliability of the smartphone-based markerless system. Movements were recorded simultaneously with two iPhones using Uplift Labs computer vision software and with Qualisys, a 12-camera marker-based motion capture system. Each participant performed two evaluations, one week apart, consisting of two countermovement jumps. Nested Bland-Altman limits of agreement (LOA), mixed-effect linear regressions, and intraclass correlation coefficients (ICC) were calculated. Twenty participants were included [Age: 24.7 (6.6) years. Height: 178.3 (4.3) cm. Mass: 86.5 (12.4) kg. Dominant Arm (Right): 17 (85%)]. Concurrent validity (LOA: 11.8 (4.4, 19.1)) demonstrated different results compared to convergence validity (Beta: 0.87 (0.68, 1.0)) and test-retest reliability (ICC: Uplift: 95.5 (90.9, 97.8). Qualisys: 94.4 (88.5, 97.3)). Uplift demonstrated greater than the a priori-determined limits of agreement across the CMJ. However, convergence validity was acceptable. Reliability suggests Uplift could be useful for tracking performance within individual athlete sessions.