Assistant Professor Wayne State University Detroit, Michigan, United States
Introduction:
Introduction: Gait impairments after stroke are caused, in part, due to altered joint mechanics. Notably, stroke survivors often exhibit paresis, spasticity, and hypertonia. Together, these symptoms can work to alter the stiffness of the joints. Altered joint stiffness can result in inefficient walking patterns; however, stiffness is rarely measured or monitored clinically. This is due to the high costs and setup times for systems used to perform these measurements (i.e., marker-based motion capture and custom robots). Therefore, this study aimed to determine if knee quasi-stiffness can be reliably measured using low-cost two-dimensional (2D) biomechanics data when compared with an expensive three-dimensional (3D) system.
Materials and
Methods:
Methods: Motion capture data were collected on seven stroke survivors while walking overground. Inverse kinematics and dynamics were performed using musculoskeletal modeling software (OpenSim) to obtain 3D joint kinematics and kinetics. Joint centers were tracked during each trial, and the point kinematics were dimensionally reduced by removing the medial/lateral component, creating a set of 2D data. Custom MATLAB scripts were used to calculate the quasi-stiffness of the knee joint by plotting joint torque against the joint angle. The linear portions of the plot were selected and analyzed to determine knee stiffness from the 3D and 2D data.
Results, Conclusions, and Discussions:
Results: Each walking trial had two regions of interest for the knee throughout the gait cycle, including the stiffness after initial contact (IC) and during weight acceptance (WA). Hence, there were two stiffness measurements per walk trial. The average IC stiffness calculated from 3D data was 4.92 ± 4.17 Nm/deg (mean ± SD), while the average IC stiffness calculated from 2D data was 3.73 ± 1.47 Nm/deg. The average WA stiffness calculated from 3D data was 3.89 ± 2.39 Nm/deg, while the average WA stiffness calculated from 2D data was 3.83 ± 2.22 Nm/deg. For both the IC and WA stiffness, the 3D and 2D data showed a strong correlation (r2>0.959).
Discussion: This study is the first to characterize quasi-stiffness of the knee joint during walking in chronic stroke survivors, where stiffness greatly differs from uninjured individuals. Notably, the quasi-stiffness measurements calculated from 3D data and 2D showed a strong correlation. This suggests that low-cost 2D systems (e.g., webcam-based) could potentially be used to allow clinics to measure joint mechanics during walking, a metric that has not been achievable previously. This current study provides an initial characterization and validation of this low-cost technique for measuring quasi-stiffness of the knee joint in chronic stroke survivors. Further analysis is ongoing, looking at adding additional stroke survivors and other joints (i.e., hip and ankle).
Conclusion: This study not only provides a characterization of the quasi-stiffness of the knee joint during gait in chronic stroke survivors but also offers a novel insight into how joint stiffness measurements are obtained from 3D data compared to measurements from 2D data. The current results, which suggest a strong correlation between 2D and 3D quasi-stiffness measurements, have the potential to significantly improve current rehabilitative metrics.
