Effect of hip abduction assistance on metabolic cost and balance during human walking.

The use of wearable robots to provide walking assistance has rapidly grown over the past decade, with notable advances made in robot design and control methods toward reducing physical effort while performing an activity. The reduction in walking effort has mainly been achieved by assisting forward progression in the sagittal plane. Human gait, however, is a complex movement that combines motions in three planes, not only the sagittal but also the transverse and frontal planes. In the frontal plane, the hip joint plays a key role in gait, including balance. However, wearable robots targeting this motion have rarely been investigated. In this study, we developed a hip abduction assistance wearable robot by formulating the hypothesis that assistance that mimics the biological hip abduction moment or power could reduce the metabolic cost of walking and affect the dynamic balance. We found that hip abduction assistance with a biological moment second peak mimic profile reduced the metabolic cost of walking by 11.6% compared with the normal walking condition. The assistance also influenced balance-related parameters, including the margin of stability. Hip abduction assistance influenced the center-of-mass movement in the mediolateral direction. When the robot assistance was applied as the center of mass moved toward the opposite leg, the assistance replaced some of the efforts that would have otherwise been provided by the human. This indicates that hip abduction assistance can reduce physical effort during human walking while influencing balance.

Deep Domain Adaptation, Pseudo-Labeling, and Shallow Network for Accurate and Fast Gait Prediction of Unlabeled Datasets.

Developing personalized gait phase prediction models is difficult because acquiring accurate gait phases requires expensive experiments. This problem can be addressed via semi-supervised domain adaptation (DA), which minimizes the discrepancy between the source and target subject features. However, classical DA models have a trade-off between accuracy and inference speed. Whereas deep DA models provide accurate prediction results with a slow inference speed, shallow DA models produce less accurate results with a fast inference speed. To achieve both high accuracy and fast inference, a dual-stage DA framework is proposed in this study. The first stage uses a deep network for precise DA. Then, a pseudo-gait-phase label of the target subject is obtained using the first-stage model. In the second stage, a shallow but fast network is trained using the pseudo-label. Because computation for DA is not conducted in the second stage, an accurate prediction can be accomplished even with the shallow network. Test results show that the proposed DA framework reduces the prediction error by 1.04% compared with a shallow DA model while maintaining its fast inference speed. The proposed DA framework can be used to provide fast personalized gait prediction models for real-time control systems such as wearable robots.

Development of a maxillomandibular arch form based on the center of resistance of teeth using cone-beam computed tomography.

INTRODUCTION: Controlling transverse discrepancies is necessary to ensure stable and functional occlusion. Altered molar inclinations can camouflage the transverse relationship. The purpose of this research was to evaluate the maxillomandibular relationship of the center of resistance (CR) of the arch form created by the CR of teeth and compare these CR arch forms by their skeletal patterns. METHODS: Sixty patients with minor crowding and normal posterior overjet were divided into 3 groups according to ANB angle: skeletal Class I group had ANB angle between 0° and 4° (n = 20), skeletal Class II group had ANB angle >4° (n = 20), and skeletal Class III group had ANB angle <0° (n = 20). The 3-dimensional coordinates of the CR were estimated using cone-beam computed tomography images and projected on the CR occlusal plane to obtain the 2-dimensional coordinates. The CR arch forms were constructed and evaluated using Matlab (MathWorks, Natick, Mass). RESULTS: On comparing maxillomandibular CR arch form widths, the maxilla was significantly larger than the mandible of the canine and first premolar. The mandible was larger in the first molar of the skeletal Class III group. The maxillomandibular CR arch form width ratios were between 0.97 and 1.35. On comparing maxillomandibular CR arch form areas, the maxilla was significantly larger than the mandible in the anterior segment, and the mandible was larger in the posterior segment. The ratios were between 0.86 and 2.25. In between-group comparison, the skeletal Class III group showed significantly greater arch forms in the mandible. CONCLUSIONS: CR arch forms had significant maxillomandibular differences throughout the arch. The maxillomandibular ratios could be a reference for site-specific transverse discrepancy analysis.

Unsupervised Gait Phase Estimation With Domain-Adversarial Neural Network and Adaptive Window.

The performanceof previous machine learning models for gait phase is only satisfactory under limited conditions. First, they produce accurate estimations only when the ground truth of the gait phase (of the target subject) is known. In contrast, when the ground truth of a target subject is not used to train an algorithm, the estimation error noticeably increases. Expensive equipment is required to precisely measure the ground truth of the gait phase. Thus, previous methods have practical shortcoming when they are optimized for individual users. To address this problem, this study introduces an unsupervised domain adaptation technique for estimation without the true gait phase of the target subject. Specifically, a domain-adversarial neural network was modified to perform regression on continuous gait phases. Second, the accuracy of previous models can be degraded by variations in stride time. To address this problem, this study developed an adaptive window method that actively considers changes in stride time. This model considerably reduces estimation errors for walking and running motions. Finally, this study proposed a new method to select the optimal source subject (among several subjects) by defining the similarity between sequential embedding features.

Collective intracellular cargo transport by multiple kinesins on multiple microtubules.

The transport of intracellular organelles is accomplished by groups of molecular motors, such as kinesin, myosin, and dynein. Previous studies have demonstrated that the cooperation between kinesins on a track is beneficial for long transport. However, within crowded three-dimensional (3D) cytoskeletal networks, surplus motors could impair transport and lead to traffic jams of cargos. Comprehensive understanding of the effects of the interactions among molecular motors, cargo, and tracks on the 3D cargo transport dynamics is still lack. In this work, a 3D stochastic multiphysics model is introduced to study the synergistic and antagonistic motions of kinesin motors walking on multiple mircotubules (MTs). Based on the model, we show that kinesins attaching to a common cargo can interact mechanically through the transient forces in their cargo linkers. Under different environmental conditions, such as different MT topologies and kinesin concentrations, the transient forces in the kinesins, the stepping frequency and the binding and unbinding probabilities of kinesins are changed substantially. Therefore, the macroscopic transport properties, specifically the stall force of the cargo, the transport direction at track intersections, and the mean-square displacement (MSD) of the cargo along the MT bundles vary over the environmental conditions. In general, conditions that improve the synergistic motion of kinesins increase the stall force of the cargo and the capability of maintaining the transport. In contrast, the antagonistic motion of kinesins temporarily traps the cargo and slows down the transport. Furthermore, this study predicts an optimal number of kinesins for the cargo transport at MT intersections and along MT bundles.

Dynamic model for kinesin-mediated long-range transport and its local traffic jam caused by tau proteins.

In neurons, several intracellular cargoes are transported by motor proteins (kinesins) which walk on microtubules (MTs). However, kinesins can possibly unbind from the MTs before they reach their destinations. The unbound kinesins randomly diffuse in neurons until they bind to MTs. Then, they walk again along the MTs to continue their tasks. Kinesins repeat this cycle of motion until they transport their cargoes to the destinations. However, most previous models mainly focused on the motion of kinesins when they walk on MTs. Thus, a new model is required to encompass the various types of kinesin motion. We developed a comprehensive model and studied the long-range axonal transport of neurons using the model. To enhance reliability of the model, it was constructed based on multiphysics on kinesin motion (i.e., chemical kinetics, diffusion, fluid dynamics, nonlinear dynamics, and stochastic characteristics). Also, parameter values for kinesin motions are carefully obtained by comparing the model predictions and several experimental observations. The axonal transport can be degraded when a large number of binding sites on MTs are blocked by excessive tau proteins. By considering the interference between walking kinesins and tau molecules on MTs, effects of tau proteins on the axonal transport are studied. One of the meaningful predictions obtained from the model is that the velocity is not an effective metric to estimate the degradation of the transport because the decrease in velocity is not noticeable when the concentration of tau protein is not high. However, our model shows that the transport locally changes near tau molecules on MTs even when the change in the velocity is not significant. Thus, a statistical method is proposed to detect this local change effectively. The advantage of this method is that a value obtained from this method is highly sensitive to the concentration of tau protein. Another benefit of this method is that this highly sensitive value can be acquired with relatively low precision and low temporal resolution considering the time scale and length scale of the kinesin motion. This method can be used to estimate the condition of the axonal transport system.

Effects of Obstacles on the Dynamics of Kinesins, Including Velocity and Run Length, Predicted by a Model of Two Dimensional Motion.

Kinesins are molecular motors which walk along microtubules by moving their heads to different binding sites. The motion of kinesin is realized by a conformational change in the structure of the kinesin molecule and by a diffusion of one of its two heads. In this study, a novel model is developed to account for the 2D diffusion of kinesin heads to several neighboring binding sites (near the surface of microtubules). To determine the direction of the next step of a kinesin molecule, this model considers the extension in the neck linkers of kinesin and the dynamic behavior of the coiled-coil structure of the kinesin neck. Also, the mechanical interference between kinesins and obstacles anchored on the microtubules is characterized. The model predicts that both the kinesin velocity and run length (i.e., the walking distance before detaching from the microtubule) are reduced by static obstacles. The run length is decreased more significantly by static obstacles than the velocity. Moreover, our model is able to predict the motion of kinesin when other (several) motors also move along the same microtubule. Furthermore, it suggests that the effect of mechanical interaction/interference between motors is much weaker than the effect of static obstacles. Our newly developed model can be used to address unanswered questions regarding degraded transport caused by the presence of excessive tau proteins on microtubules.

Highly loaded behavior of kinesins increases the robustness of transport under high resisting loads.

Kinesins are nano-sized biological motors which walk by repeating a mechanochemical cycle. A single kinesin molecule is able to transport its cargo about 1 µm in the absence of external loads. However, kinesins perform much longer range transport in cells by working collectively. This long range of transport by a team of kinesins is surprising because the motion of the cargo in cells can be hindered by other particles. To reveal how the kinesins are able to accomplish their tasks of transport in harsh intracellular circumstances, stochastic studies on the kinesin motion are performed by considering the binding and unbinding of kinesins to microtubules and their dependence on the force acting on kinesin molecules. The unbinding probabilities corresponding to each mechanochemical state of kinesin are modeled. The statistical characterization of the instants and locations of binding are captured by computing the probability of unbound kinesin being at given locations. It is predicted that a group of kinesins has a more efficient transport than a single kinesin from the perspective of velocity and run length. Particularly, when large loads are applied, the leading kinesin remains bound to the microtubule for long time which increases the chances of the other kinesins to bind to the microtubule. To predict effects of this behavior of the leading kinesin under large loads on the collective transport, the motion of the cargo is studied when the cargo confronts obstacles. The result suggests that the behavior of kinesins under large loads prevents the early termination of the transport which can be caused by the interference with the static or moving obstacles.

Metrics for characterizing collective transport by multiple dimeric kinesins.

Kinesin is a processive molecular motor which transports various cellular cargos by converting chemical energy into mechanical movements. Although the motion of a single molecule has been characterized in several studies, the dynamics of collective transport remains controversial. Since the chemical reactions fueling molecular motors are stochastic processes, the movements of coupled motors are not perfectly synchronized. The goal of this study is to develop metrics to analyze the level of synchronization of coupled (stochastic) motors. The correlation among movements of coupled motors, the slackness, the cooperativity, and the power loss of kinesins are explored using the developed metrics. These metrics can be extended to characterize collective work done by other molecular motors also.

The effects of viscoelastic fluid on kinesin transport.

Kinesins are molecular motors which transport various cargoes in the cytoplasm of cells and are involved in cell division. Previous models for kinesins have only targeted their in vitro motion. Thus, their applicability is limited to kinesin moving in a fluid with low viscosity. However, highly viscoelastic fluids have considerable effects on the movement of kinesin. For example, the high viscosity modifies the relation between the load and the speed of kinesin. While the velocity of kinesin has a nonlinear dependence with respect to the load in environments with low viscosity, highly viscous forces change that behavior. Also, the elastic nature of the fluid changes the velocity of kinesin. The new mechanistic model described in this paper considers the viscoelasticity of the fluid using subdiffusion. The approach is based on a generalized Langevin equation and fractional Brownian motion. Results show that a single kinesin has a maximum velocity when the ratio between the viscosity and elasticity is about 0.5. Additionally, the new model is able to capture the transient dynamics, which allows the prediction of the motion of kinesin under time varying loads.