Accuracy and reliability of the Memsens system to evaluate a squat jump

International audienceDuring these last five years, devices with small accelerometer have been increasingly usedto measure human sport motion. Few studies have defined the measurement errors andlimits of agreement of Chose sensors in the particular case of squat jumps when theaccelerometer was attached on the athlete.The aim of this study is to define the accuracy and reliability of measurement with the useof a nIMU Memsens sensor. Measurements of Memsens device is compared to these of aforce plate. Nine male subjects (mean f standard deviation), with 179.8 ± 5.12 cm height and76.02 ± 6.9 kg body mass performed squat jumps on a force plate (Kisler, 500 Hz). For eachjump, the nIMU sensor (Memsens, USA) was attached at the hip of the athlete. TheMemsens device includes the nIMU (120 Hz) sensor and software developed at INSEPlaboratory. The device finds the motion start, defines the vertical acceleration of the subject,provides a 500 Hz interpolation, and then estimates the maximal velocity (Vmax), the takeoff velocity (Vtoff) and the flight time (t). The data of both devices of measurement(force plate and Memsens system) are synchronized. For each jump, data of the force platewere converted to define Vmax, Vtoff and t. Paired t-test, R correlation coefficient andBland & Altman test were used to compare the validity and the limits of agreementbetween the two devices.Results show no significant difference between the measurement of Vmax, Vtoff and t. Thecorrelation between the tool’s measurements is R> 0.91 for each parameter. Bland &Altman test shows very low bias and good reliability between measurements. Inconclusion, the Memsens system can be used to evaluate the squat jump with acceptableaccuracy and reliability. This study shows that a sensor attached on the hip permits toestimate kinetic’s variables of a squat jump in agreement with measurement of the centre ofmass maximal velocity time flight as well as using a force plate. The underestimation ofthe take off velocity using the sensor could be explained by the difference of the point ofmeasurement between the devices. In the case of training and testing, this study shows thata sensor using tri-axial accelerometers can help to estimate human motion accurately. Thisnew device could help the athletic subject to test his physical abilities without laboratoryrestriction