Sprint mechanics evaluation using inertial sensor-based technology : A laboratory validation study

Abstract

Advances in micro-electromechanical systems have turned magnetic inertial measurement units (MIMUs) into a suitable tool for vertical jumping biomechanical evaluation. Thus, this study aimed to determine whether appropriate reliability and agreement reports could also be obtained when analyzing 20-m sprint mechanics. Four bouts of 20-m sprints were evaluated to determine whether the data provided by a MIMU placed at the lumbar spine could reliably assess sprint mechanics and to examine the validity of the MIMU sensor compared to force plate recordings. Maximal power (P0), force (F0), and velocity (V0), as well as other mechanical determinants of sprint performance associated with the force-velocity, power-velocity, and ratio of forces-velocity, such as applied horizontal force loss (Sfv) and decrease in ratio of forces (Drf), were calculated and compared between instrumentations. Extremely large-to-very large correlation levels between MIMU sensor-based sprint mechanics variables and force plate recordings were obtained (mean±SD, force plate vs MIMU; V0, 8.61±0.85 vs 8.42±0.69; F0, 383±110 vs 391±103; P0, 873±246 vs 799±241; Sfv, −44.6±12.7 vs −46.2±10.7), ranging from 0.88 to 0.94, except for Drf, which showed weak-to-moderate correlation level (r=.45; −6.32±1.08 vs −5.76±0.68). Step-averaged force values measured with both systems were highly correlated (r=.88), with a regression slope close to the identity (1.01). Bland and Altman graphical representation showed a no random distribution of measured force values. Finally, very large-to-extremely large retest correlation coefficients were found for the intertrial reliability of MIMU measurements of sprint performance variables (r value ranging from .72 to .96). Therefore, MIMUs showed appropriate validity and reliability values for 20-m sprint performance variables.