Reliability quantification and gait loading asymmetry assessment with wearable insoles in transfemoral amputees

Objective Gait defaults of amputees are known to be increased in difficult daily living walking conditions. Among quantitative parameters, load asymmetry is frequently used to assess gait quality. Replication of realistic daily living environment in a motion analysis laboratory (MAL) is difficult. Wearable pressure insoles, by providing normal ground reaction force (NGrF), can be used to quantify load asymmetry in real life conditions. This indicator is useful for physicians to monitor the rehabilitation progress or the prosthetic fitting suitability. The aim of the study was to quantify the reliability of NGrF measurement with Pedoped ® insoles, which gives NGrF in real time with biofeedback, and to assess gait asymmetry in transfemoral (TF) amputees at different walking speeds with the insoles. Material/patients and methods This prospective study included 5 TF amputees. In a MAL, they walked with their own prosthetic devices at three self-selected speeds on level ground, wearing Pedoped ® insoles. Reliability was assessed by comparing NGrF obtained from both systems using Bland-Altman plots and normalized RMSE (NRMSE). Load asymmetry between prosthetic and intact limb at the three different speeds was computed by Absolute Symmetry Index (ASI) on the following parameters: first and second maxima of NGrF and stance phase duration. The mean error was also assessed for ASI parameters computed with both systems. Results NRMSE was 7.2% (± 2.8%) for the prosthetic side and 9.8% ( ± 3.5%) for the intact side. Stance phase duration for both systems was consistently similar. The Bland and Altmann plots showed over 95% of the values between the limits of agreement. Mean error of ASI parameters ranged from, 6.2% to 1.4%, with an insole overestimation. All ASI were positive showing a higher value of the parameters of the intact limb compared to the prosthetic limb. Load asymmetry significantly increased with walking speed. Discussion, conclusion The results suggested that the insoles could be used to evaluate load asymmetry using ASI parameters showing the overload on the intact limb, which is consistent with the literature. The biofeedback of NGrF in real time opens perspectives for the improvement of rehabilitation techniques. Connected auto-rehabilitation could be very useful to enhance the recovering of loading symmetry during gait.