Subjects:
Twenty-five young healthy participants (14 males, 11 females), adults aged from (20 to 35) years old, without any pathology were included in the study. The mean age of the patients was 26.7 ± 3.4 years, the mean body height was 175.8 ± 9.4 cm and the mean body weight was 74.4 ± 11.2 kg. The mean body mass index (BMI) was 24 ± 2.7 kg/m2. The following exclusion criteria were established: no previous history of either orthopedic or neurological ailments, such as a recent injury or surgery, which could affect their walking pattern. Furthermore, the right leg has to be the dominant leg.
Measurement setup:
To measure muscle activity and knee flexion, 8 sensors for surface electromyography (sEMG) and 7 inertial measurement units (IMUs) from the MyoMuscle and MyoMotion system (Noraxon U.S.A. Inc., Scottsdale, USA) were used. Four muscles from the right leg were analyzed the knee extensor (vastus medialis), the knee flexor (biceps femoris), the plantar flexor (gastrocnemius medialis), and the dorsiflexor (anterior tibialis). Before the electrodes and corresponding sensors were attached, hair was shaved, and skin was disinfected. The EMG electrodes (Noraxon´s disposable self-adhesive dual Ag/AgCl electrodes) were positioned according to the Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) recommendations (Fig. 1), longitudinally over the muscle bellies of the muscles outlined above [28]. The seven inertial sensors were attached to the body with elastic straps or tape based on the guideline instructions, the reference sensor was attached to the sacrum (pelvis). The thigh sensors were attached with elastic straps or tape on the middle of the thigh, the same methods were used with the shank sensors which were attached slightly to the medial side of the tibia, Furthermore, the feet sensors were attached by a clip in the top of the shoes (Fig. 1). The aim of using the IMU system was to measure knee flexion during the different trials.
Figure 1: EMG and IMU sensors position with the brace
All participants wore the same knee osteoarthritis brace (Valgus and varus) M.4s OA 3-points pressure valgus - varus osteoarthritis knee brace and the 4-points principle of stabilization (Medi GmbH & Co. KG, Bayreuth, Germany) (Fig. 2). All kinematic measurements were recorded with a sampling frequency of 100 Hz [29].
Figure 2:The M.4s OA comfort. The brace can be Adjusted in case of Osteoarthritis
Test procedure:
The measurements took place in the RWTH Aachen University Hospital, Germany. Each measurement lasted approximately one hour and a half. Before starting with the measurements, the participants were asked to perform the calibration procedure for the IMU sensor system, where the participants had to hold a static standing in the upright position (neutral zero joint position). After calibration, the participants were asked to perform the two following tests: 10-meter walking test (10MWT), and stair climbing (STC). To minimize the number of external influences and boundary conditions participants were asked to maintain their walking speed instead of a default speed. Hence, the only difference in the next step was the addition of the brace. The 10MWT was conducted in a hallway in the hospital. Regarding the STC, the participants performed on seven stairs with dimensions of 26 x 84 x 19 cm. The study analyzed the stance and swing phases in 10MWT and STC. Each test was performed with 7 trails with different levels: without the brace (WOB), valgus brace without adjustment (Lv.0), valgus brace with adjustment level 1 (LV.1), valgus brace with adjustment level 2 (Lv.2), and the same order for varus adjustments. The two levels were identified with the rotation of the hinges above and below the knee joints in the braces; level one was identified with ¾ of the rotation and level 2 was identified with a full rotation of the hinges on both braces (29). To avoid bias in repeated measurements, with each participant the sequence of the trails was changed.
Data processing:
The measured data were systematically processed, and relevant parameters were extracted. The MyoMuscle (Clinical DTS version) measure with an initial sampling rate of 3000Hz transmits pre-processed smoothed and rectified EMG data (100ms RMS), shown as a unipolar positive signal (30). A bandpass filter was used (10-500Hz) with a 100Hz wireless update rate to the receiver. For this data, we used the Peak value (1000ms) for amplitude normalization to be able to compare the measurements with different levels and on both sides (31). By using peak value, the EMG data (measured in µV) were normalized to % of the mean muscle activity.
The MyoResearch Software (version MR 3.12, Noraxon U.S.A. Inc., Scottsdale, USA) provides the “Contact Mode” option, which can automatically detect the initial and terminal contact of each foot based on the acceleration data of the IMUs placed on the feet during walking and stair climbing (32). The muscle activity and anatomical angles regarding the gait cycle can be analyzed from this option. The continuously measured data (normalized muscle activity and anatomic angle) were split into strides, and for each stride, the x-axis (time-axis) was normalized to the gait cycle (in percent). The data of each gait cycle were interpolated to 100 points and the mean over 2 strides was calculated for all measured variables. The mean muscle activities and joint angles over the gait cycle of all participants for the different test conditions were presented with the help of MATLAB (MathWorks R2019a, MathWorks, Natick, MA, USA).
Statistical analysis:
Statistical analyses were performed using the software IBM® SPSS Statistics (IBM® SPSS Statistics v. 25, IBM Cooperation). The normal distribution of the dependent variables was examined using the Shapiro-Wilks test and was not confirmed for most variables. Therefore, nonparametric tests were used. To evaluate the effect of wearing the brace, the Wilcoxon test was used to compare results between WOB and LV0 measurements. The effect size r is equal to (Eq. 1)
\(r=\frac{z}{\sqrt{N}}\) | Equation 1 |
(with z-value from the associated Wilcoxon test and N as the number of pairs). It is presented as the coefficient of determination r², explaining the collective variance of the analyzed variables. To compare the different levels of the brace, the Friedman test was calculated with pairwise Post Hoc analysis and the brace adjustments as the independent variables. There were three groups: Lv0, Lv1, and Lv2 brace adjustment. The pairwise comparison was prepared only in case of significant results. Statistically significant differences were set at p < 0.05.