Calibration of an instrumented treadmill using a precision-controlled device with artificial neural network-based error corrections

Centre of Pressure Estimation during Walking Using Only Inertial-Measurement Units and End-To-End Statistical Modelling

Estimation of the centre of pressure (COP) is an important part of the gait analysis, for example, when evaluating the functional capacity of individuals affected by motor impairment. Inertial measurement units (IMUs) and force sensors are commonly used to measure gait characteristic of healthy and impaired subjects. We present a methodology for estimating the COP solely from raw gyroscope, accelerometer, and magnetometer data from IMUs using statistical modelling. We demonstrate the viability of the method using an example of two models: a linear model and a non-linear Long-Short-Term Memory (LSTM) neural network model. Models were trained on the COP ground truth data measured using an instrumented treadmill and achieved the average intra-subject root mean square (RMS) error between estimated and ground truth COP of 12.3 mm and the average inter-subject RMS error of 23.7 mm which is comparable or better than similar studies so far. We show that the calibration procedure in the instrumented treadmill can be as short as a couple of minutes without the decrease in our model performance. We also show that the magnetic component of the recorded IMU signal, which is most sensitive to environmental changes, can be safely dropped without a significant decrease in model performance. Finally, we show that the number of IMUs can be reduced to five without deterioration in the model performance.

Comparison of body's center of mass motion relative to center of pressure between treadmill and over-ground walking

Treadmills have been used in rehabilitation settings to provide convenient protocols and continuous monitoring of movement over multiple cycles at well-controlled speeds for gait and balance training. However, the potential differences in the movement control may affect the translation of the training outcomes to real life over-ground walking (OW). The similarities and differences in the balance control between treadmill walking (TW) and OW have largely been unexplored. The current study bridged the gap by comparing the motions of the body's center of mass (COM) relative to the center of pressure (COP) between TW and OW, in terms of the COM-COP inclination angle (IA) and its rate of change (RCIA). The movement of the COM and COP separately were quite different between OW and TW, but when describing the COM motion relative to the COP, the COM motions became similar qualitatively with similar butterfly patterns. However, significantly increased peak values in themediolateral RCIA and greater ranges of mediolateral IA were found during TW (p<0.004). In the sagittal plane, the posterior velocity of the belt led to an anterior RCIA (posterior RCIA in OW) with increasing anterior IA during early double-limb support phase, and reduced posterior RCIA (p<0.009) with an increased anterior IA (p<0.001) during the remainder of the phase. These differences between TW and OW may have to be taken into account in future designs of strategies to optimize the translation of treadmill gait training outcomes into real life over-ground walking.

Effects of belt speed on the body's center of mass motion relative to the center of pressure during treadmill walking

Treadmills are often used in clinical settings to improve walking balance control in patients with gait impairments. However, knowledge of the effects of belt speed on balance control remains incomplete. The current study determined such effects in terms of inclination angles (IA) and the rate of change (RCIA) of the center of mass (COM) motion relative to the center of pressure (COP) in twelve healthy adults at five belt speeds, including the subjects' preferred walking speed (PWS), as measured using a motion capture system and an instrumented treadmill. The values of IAs and RCIAs at key gait events and their average values over single-limb support (DLS) and double-limb support (DLS) were compared between speeds using one-way repeated measures analysis of variances. While the COM-COP controls were different between SLS and DLS, they were inter-related to form an integrated whole. Among the belt speeds, the range of frontal IA during SLS was smallest at the PWS (p<0.05). With increasing speed, most variables of the sagittal IAs and RCIAs, and of the frontal RCIAs during DLS showed a linearly increasing trend (p<0.001). A linearly decreasing trend was found in the frontal IA at toe-off and in the average frontal RCIA during SLS (p<0.05). The PWS appeared to be the best compromise between frontal stability during SLS and smooth weight-transfer during DLS. The current results provide useful baseline data for selecting speeds according to training needs, and may be helpful for developing protocols for gait retraining for patients with gait impairment.