Review
Copyright ©The Author(s) 2021.
World J Orthop. Jul 18, 2021; 12(7): 467-484
Published online Jul 18, 2021. doi: 10.5312/wjo.v12.i7.467
Table 1 Studies investigating the reliability of inertial measurement unit sensors
Ref.
Sensor/position
Comparison system
Results
Outcomes
Qiu et al[29], 20163 magnetic angular rate and gravity/thigh, shank, and footViconPosition accuracy of 0.3%, the ΔXY radial distance error of 0.82% and the distance error of 0.27%, position error of 0.4%The combination of distributed wearable sensors with the Denavit–Hartenberg convention resulted in a promising tool for tracking lower limb movements
Sprager et al[30], 20151 multi-sensor platform integrated into a smart garment/kneeNPGood activity discrimination can be achieved based RMSE and SD from flexible sensor, acceleration and gyroscope dataPreliminary results show that walking, running, stairs climbing can be discriminated based on the data collected
Cresswell et al[33], 20174 Shimmer3 sensor nodes/all sides of the shankNPThe results of the fixed effects models highlighted the discrepancies between front–back mounting versus inner–outer mountingFor y-axis gyroscope data, the variation is mostly influenced by mounting location. Mounting location should not vary but if it has to vary, it is better for it to vary between inner and outer leg mounting locations
Fusca et al[35], 20181 IMU/posterior CoMElite (BTS)Mean absolute percentage error of: Stride time is 5.7%; Cadence is 4.9%; Step's length is 5.6%; Step's speed is 13.5%The use of IMU at CoM presents a good reliability for carrying out ambulatory, long-term, and ecologic kinematic of gait analysis
Saggio et al[36], 20207 IMU/pelvis, thighs, shanks and feetViconJoints ROMs RMSE and ICC PCC > 0.75, Reliability all the ICC > 0.975IMUs sensors showed a high reliability on joints' movement and walking test