Walking speed is a confounding factor in biomechanical analyses of gait, but still many studies compare gait biomechanics at comfortable walking speed (CWS) that is likely to differ between groups or conditions. To identify gait deviation unrelated to walking speed, methods are needed to correct biomechanical data over the gait cycle for walking speed. How to compare knee kinetics over the gait cycle at different walking speeds? 22 asymptomatic subjects walked on a dual-belt treadmill at CWS and 4 fixed speeds. Knee moments in sagittal (KFM) and frontal plane (KAM) were calculated via inverse dynamics. The net moment differences between CWS and fixed speed were expressed as a root-mean-square error (RMSE) normalized to the range of the variable. Two methods to correct for walking speed were compared. In method 1, KFM and KAM values were estimated based on interpolation between speeds at each percentage of the gait cycle. In method 2, principal component analysis was used to extract speed related features to reconstruct KFM and KAM at the speed of interest. The accuracy of both methods was tested using a leave-one-out cross validation. Walking speed influenced the magnitude and shape of KFM and KAM. To account for these speed influences using both methods, leave-one-out cross validation showed low normalized RMSE (< 5 %), with little difference between the two methods. RMSE for both reconstruction methods were up to 60 % lower than the RMSE between CWS and fixed speed. Both methods could accurately correct knee kinetics over the gait cycle for the effects of walking speed. Walking speed dependency should be incorporated in each gait laboratory's reference dataset to be able to identify gait deviations unrelated to gait speed.
24 September 2021 in IEEE Transactions on Biomedical Engineering
Objective : This study determined the contribution of the deep hip muscles to hip stability. Methods : Hip stability was defined as rotational hip stiffness in the sagittal plane, which was calculated for walking trials for 12 participants via an electromyography (EMG)–informed neuromusculoskeletal model which included all 22 hip spanning muscles. Three model configurations which differed in deep hip muscle excitations but had identical excitations for all other muscles were compared: (1) deep hip muscles informed by intramuscular EMG measurements ( assisted activation ); (2) deep hip muscles with simulated zero activation ( no activation ); (3) deep hip muscles with simulated maximal activation ( maximal activation ). Sagittal plane rotational hip stiffness across the gait cycle was compared between the three model configurations using a within-participant analysis of variance via statistical parametric mapping (p < 0.05). Results : Compared to the assisted activation configuration, hip stiffness (mean (95% confidence interval)) was 0.8% (0.7 to 0.9) lower in the no activation configuration, and 3.2% (3.0 to 3.4) higher in the maximal activation configuration. Conclusion : Regardless of activation level, deep hip muscles contributed little to sagittal plane rotational hip stiffness, which casts doubt on their assumed function as hip stabilizers. Significance : The merit of targeted deep hip muscle strengthening to improve hip stability in rehabilitation programs remains unclear.
Objective: Markerless motion tracking using computer vision offers a scalable alternative to traditional marker-based analysis. While insight on how different emerging solutions compare could inform adoption across applications by highlighting accuracy-complexity tradeoffs, comprehensive benchmarking of these methods on the same open dataset remains a gap. This study evaluated thirteen single-view and two multi-view markerless motion capture methods against marker-based motion tracking for lower-extremity kinematics. Twenty-three healthy adults performed walking and lower-limb functional exercises, recorded with 20 infrared and 10 red-green-blue (RGB) cameras. The best-performing single-view model (WHAM) was used to test the hypothesis that the anatomical constraints imposed by biomechanical models improve the accuracy of vision-only solutions. We found that the addition of biomechanical modeling (bioWHAM) does not significantly improve overall kinematics accuracy, with root-mean-square differences (RMSD) sometimes being lower for the baseline model (WHAM) and sometimes for bioWHAM, with the median fluctuations across tasks being under 1.7$^\circ $, in either direction. Multi-view methods generally outperformed single-view ones: OpenCap, an open-source solution using two cameras, outperformed WHAM by 1.7$^\circ $ (p < 0.0001), whereas Theia3D, a commercial software using ten cameras, outperformed OpenCap by 1.3$^\circ $ (p < 0.0001). These findings suggest that while multi-view systems can enhance accuracy, the marginal benefits may not justify the added complexity across all applications. In addition to helping users consider the hardware, software, and accuracy tradeoffs, these findings highlight the need for continued innovation in multi-view fusion and incorporation of biomechanical modeling and computer vision. The accompanying dataset (I-MOVE-23) should facilitate continued benchmarking of emerging solutions.