Lower-Limb Myoelectric Calibration Postures for Transtibial Prostheses.

The use of an agonist-antagonist muscle pair for myoelectric control of a transtibial prosthesis requires normalizing the myoelectric signals and identifying their co-contraction signature. Extensive literature has explored the relationship between body posture and lower-limb muscle activation level using surface electromyography (EMG), but it is unknown how these relationships hold after amputation. Using a virtual tracking task, this study compares the effect of three different calibration postures (seated, standing, dynamic) on user tracking ability while in two tracking postures (seated, standing) for 18 able-bodied (AB) subjects and 9 subjects with transtibial (TT) amputation. As expected, AB subjects produced statistically significant differences in muscle activation for gastrocnemius (GAS) when seated vs. standing during calibration (p = 8.8e-4), but not for tibialis anterior (TA) (p = 0.76). TT subjects, however, showed no significant differences in GAS or TA between seated and standing (p = 0.90, 0.60 respectively). It was also determined that normalizing EMG by the global maximum signal observed (standard in biomechanic analysis) is undesirable for myoelectric control. For best general results with this framework, calibration in both seated and dynamic postures is recommended, taking the normalization information from the seated posture and the narrowest co-contraction slopes from the two.

Finite-State Impedance and Direct Myoelectric Control for Robotic Ankle Prostheses: Comparing Their Performance and Exploring Their Combination.

Non-volitional control, such as finite-state machine (FSM) impedance control, does not directly incorporate user intent signals, while volitional control, like direct myoelectric control (DMC), relies on these signals. This paper compares the performance, capabilities, and perception of FSM impedance control to DMC on a robotic prosthesis for subjects with and without transtibial amputation. It then explores, using the same metrics, the feasibility and performance of the combination of FSM impedance control and DMC across the full gait cycle, termed Hybrid Volitional Control (HVC). After calibration and acclimation with each controller, subjects walked for two minutes, explored the control capabilities, and completed a questionnaire. FSM impedance control produced larger average peak torque (1.15 Nm/kg) and power (2.05 W/kg) than DMC (0.88 Nm/kg and 0.94 W/kg). The discrete FSM, however, caused non-standard kinetic and kinematic trajectories, while DMC yielded trajectories qualitatively more similar to able-bodied biomechanics. While walking with HVC, all subjects successfully achieved ankle push-off and were able to modulate the magnitude of push-off via the volitional input. Unexpectedly, however, HVC behaved either more similarly to FSM impedance control or to DMC alone, rather than in combination. Both DMC and HVC, but not FSM impedance control, allowed subjects to achieve unique activities such as tip-toe standing, foot tapping, side-stepping, and backward walking. Able-bodied subject (N=6) preferences were split amongst the controllers, while all transtibial subjects (N=3) preferred DMC. Desired performance and ease of use showed the highest correlations with overall satisfaction (0.81 and 0.82, respectively).

Data-efficient human walking speed intent identification.

The ability to accurately identify human gait intent is a challenge relevant to the success of many applications in robotics, including, but not limited to, assistive devices. Most existing intent identification approaches, however, are either sensor-specific or use a pattern-recognition approach that requires large amounts of training data. This paper introduces a real-time walking speed intent identification algorithm based on the Mahalanobis distance that requires minimal training data. This data efficiency is enabled by making the simplifying assumption that each time step of walking data is independent of all other time steps. The accuracy of the algorithm was analyzed through human-subject experiments that were conducted using controlled walking speed changes on a treadmill. Experimental results confirm that the model used for intent identification converges quickly (within 5 min of training data). On average, the algorithm successfully detected the change in desired walking speed within one gait cycle and had a maximum of 87% accuracy at responding with the correct intent category of speed up, slow down, or no change. The findings also show that the accuracy of the algorithm improves with the magnitude of the speed change, while speed increases were more easily detected than speed decreases.

Characterizing Intent Changes in Exoskeleton-Assisted Walking Through Onboard Sensors.

Robotic exoskeletons are a promising technology for rehabilitation and locomotion following musculoskeletal injury, but their adoption outside the physical therapy clinic has been limited by relatively primitive methods for identifying and incorporating the user's gait intentions. Various intent detection approaches have been demonstrated using electromyography and electroencephalography signals. These technologies sense the human directly but introduce complications for donning/doffing the device and in measurement consistency. By contrast, sensors onboard the exoskeleton avoid these complications but sense the human indirectly via the human-robot interface. This pilot study examines if onboard sensors alone may enable identification of user intent. Joint positions and commanded motor currents are compared prior to and after changes in the user's intended gait speed. Preliminary experimental results confirm that these measures are significantly different following intent changes for both able-bodied and non-able-bodied users. The findings suggest that intent detection is possible with onboard sensors alone, but the intent signals depend on exoskeleton control settings, user ability, and temporal considerations.

Effects of Filtering the Center of Pressure Feedback Provided in Visually Guided Mediolateral Weight Shifting.

Thirty healthy adults completed a mediolateral weight-shifting balance task in which they were instructed to shift their weight to visually displayed target regions. A model-based filter and three different moving average filters employing 10, 34, and 58 samples were applied to the center of pressure visual feedback that guided the activity. The effects of filter selection on both the displayed feedback and the shift performance were examined in terms of shift time and non-minimum phase behavior. Shift time relates to feedback delay and shift speed, whereas non-minimum phase behavior relates to the force applied in shift initiation. Results indicated that increasing the number of samples in moving average filters (indicative of stronger filtering) significantly increases shift speed and shift initiation force. These effects indicate that careful selection and documentation of data filtering is warranted in future work and suggest opportunities for strategic filtering of visual feedback in clinical weight-shifting balance activities in order to improve outcomes based on such feedback.

Training conditions that best reproduce the joint powers of unsupported walking.

OBJECTIVE: To identify the clinically relevant combinations of body weight support and speed that best reproduce the joint powers of unsupported walking. METHODS: Timing and magnitude of lower extremity joint powers were calculated for 8 neurologically intact volunteers (4M/4F) walking with 0%, 30% and 50% body weight support at three speeds (slow, comfortable, and fast). Lower extremity joint power absorption was analyzed during weight acceptance and forward propulsion. In addition, power generation was analyzed during forward propulsion. Timings and magnitudes of joint powers per condition were evaluated to identify the training combinations of body weight support and speed that best preserved the powers of unsupported walking at slow, comfortable and fast speeds. RESULTS: For all speeds examined, increasing body weight support to 30% without changing speed provided the best match. In general, changes in speed disrupted the joint power magnitudes and timings more than application of body weight support. Increasing body weight support when faster training speeds were used proved a viable method for reproducing the joint powers of slow, unsupported walking. CONCLUSIONS: These data provide a reference for understanding the effect of potential training conditions on power absorption and generation within the lower extremity joints during walking. It is possible to reproduce the joint powers of unsupported walking with certain combinations of body weight support and speed. We recommend applying adequate levels of BWS when training speeds are faster than the overground speed goal, as occurs during treadmill-based locomotor rehabilitation of individuals with incomplete spinal cord injury.

Interpreting lateral dynamic weight shifts using a simple inverted pendulum model.

Seventy-five young, healthy adults completed a lateral weight-shifting activity in which each shifted his/her center of pressure (CoP) to visually displayed target locations with the aid of visual CoP feedback. Each subject's CoP data were modeled using a single-link inverted pendulum system with a spring-damper at the joint. This extends the simple inverted pendulum model of static balance in the sagittal plane to lateral weight-shifting balance. The model controlled pendulum angle using PD control and a ramp setpoint trajectory, and weight-shifting was characterized by both shift speed and a non-minimum phase (NMP) behavior metric. This NMP behavior metric examines the force magnitude at shift initiation and provides weight-shifting balance performance information that parallels the examination of peak ground reaction forces in gait analysis. Control parameters were optimized on a subject-by-subject basis to match balance metrics for modeled results to metric values calculated from experimental data. Overall, the model matches experimental data well (average percent error of 0.35% for shifting speed and 0.05% for NMP behavior). These results suggest that the single-link inverted pendulum model can be used effectively to capture lateral weight-shifting balance, as it has been shown to model static balance.

Predicting human walking gaits with a simple planar model.

Models of human walking with moderate complexity have the potential to accurately capture both joint kinematics and whole body energetics, thereby offering more simultaneous information than very simple models and less computational cost than very complex models. This work examines four- and six-link planar biped models with knees and rigid circular feet. The two differ in that the six-link model includes ankle joints. Stable periodic walking gaits are generated for both models using a hybrid zero dynamics-based control approach. To establish a baseline of how well the models can approximate normal human walking, gaits were optimized to match experimental human walking data, ranging in speed from very slow to very fast. The six-link model well matched the experimental step length, speed, and mean absolute power, while the four-link model did not, indicating that ankle work is a critical element in human walking models of this type. Beyond simply matching human data, the six-link model can be used in an optimization framework to predict normal human walking using a torque-squared objective function. The model well predicted experimental step length, joint motions, and mean absolute power over the full range of speeds.

A new look at an old problem: defining weight acceptance in human walking.

BACKGROUND: Weight acceptance (WA) is an important phase of bipedal gait that has received relatively little study to date. This study tested the hypothesis that the first peak knee flexion would better demarcate the end of WA power absorption activity across varying gait speeds than would the more commonly used event of contralateral toe off (CTO) or the peak hip adduction angle. METHODS: Eight control subjects (4F/4M) walked on a treadmill at slow, self-selected, and fast speeds. Kinematic and kinetic data were recorded. Joint angles and power absorption were analyzed about the, lower extremity joints (sagittal ankle, knee, hip and frontal hip). Differences in event timings and, magnitudes of negative work were analyzed (ANOVA). RESULTS: Knee sagittal power absorption continued after the CTO event at self-selected (p=0.009) and fast speeds (p=0.001), while hip frontal power absorption continued after the CTO event at slow (p=0.019), self-selected (p=0.001), and fast speeds (p=0.001). The contribution of frontal hip to overall power absorption increased as speed decreased. DISCUSSION: Peak hip adduction angle is the best kinematic marker of the end of the WA phase, and peak knee flexion angle is the best alternative marker across speeds. CTO is only appropriate to use when gait speeds are slow. In addition, the relative contribution of power absorbed in the frontal hip during WA highlights the importance of frontal plane pelvic control in locomotion, especially when gait speed is slow.

Relative efficacy of various strategies for visual feedback in standing balance activities.

Seventy-nine young, healthy adults were led through static balance and weight-shifting activities in order to study the effects of visual feedback on balance. Based on their performance, the relative effects of various feedback properties were analyzed: (1) arrangement [direct center of pressure (CoP) vs. lateral weight distribution feedback], (2) numbers (presence vs. absence of numeric feedback), and (3) dimensionality (1D vs. 2D CoP information). In the static balance activity, subjects were instructed to maintain equal weight across both feet; in the dynamic weight-shifting activity, subjects were instructed to shift their weight to each displayed target location. For static balance, lateral symmetry and sway were measured by classical parameters using CoP, center of gravity (CoG), and the difference between the two (CoP-CoG). Weight-shifting balance performance was measured using the time required to shift between target CoP positions. Results indicated that feedback arrangement had a significant effect on static sway and dynamic weight shifting, with direct CoP feedback resulting in better balance performance than lateral weight distribution. Also, numbers had a significant effect on static sway, reducing lateral sway compared to feedback without numbers. Finally, 2D CoP feedback resulted in faster performance than 1D CoP feedback in dynamic weight shifting. These results show that altering different properties of visual feedback can have significant effects on resulting balance performance; therefore, proper selection of visual feedback strategy needs to take these effects into consideration.

Invariant geometric characteristics of spatial arm motion.

This paper examines up to third-order geometric properties of wrist path and the first-order property of wrist trajectory (wrist speed) for spatial pointing movements. Previous studies report conflicting data regarding the time invariance of wrist-path shape, and most analyses are limited to the second-order geometric property (straightness, or strictly speaking, curvature). Subjects performed point-to-point reaching movements between targets whose locations ensured that the wrist paths spanned a range of lengths and lay in various portions of the arm's spatial workspace. Movement kinematics were recorded using electromagnetic sensors located on the subject's arm segments and thorax. Analysis revealed that wrist paths tend to lie in planes and to curve more as movement speed decreases. The orientation of the wrist-path plane depends on the reaching task but does not vary significantly with movement speed. The planarity of wrist paths indicates that the paths have close to zero torsion-a third-order geometric property. Wrist-speed profiles showed multiple peaks for sufficiently slow and long lasting movements, indicating deviation from the well-known, bell-shaped profile. These kinematic findings are discussed in light of various motor control theories.

The leading joint hypothesis for spatial reaching arm motions.

The leading joint hypothesis (LJH), developed for planar arm reaching, proposes that the interaction torques experienced by the proximal joint are low compared to the corresponding muscle torques. The human central nervous system could potentially ignore these interaction torques at the proximal (leading) joint with little effect on the wrist trajectory, simplifying joint-level control. This paper investigates the extension of the LJH to spatial reaching. In spatial motion, a number of terms in the governing equation (Euler's angular momentum balance) that vanish for planar movements are non-trivial, so their contributions to the joint torque must be classified as net, interaction or muscle torque. This paper applies definitions from the literature to these torque components to establish a general classification for all terms in Euler's equation. This classification is equally applicable to planar and spatial motion. Additionally, a rationale for excluding gravity torques from the torque analysis is provided. Subjects performed point-to-point reaching movements between targets whose locations ensured that the wrist paths lay in various portions of the arm's spatial workspace. Movement kinematics were recorded using electromagnetic sensors located on the subject's arm segments and thorax. The arm was modeled as a three-link kinematic chain with idealized spherical and revolute joints at the shoulder and elbow. Joint torque components were computed using inverse dynamics. Most movements were 'shoulder-led' in that the interaction torque impulse was significantly lower than the muscle torque impulse for the shoulder, but not the elbow. For the few elbow-led movements, the interaction impulse at the elbow was low, while that at the shoulder was high, and these typically involved large elbow and small shoulder displacements. These results support the LJH and extend it to spatial reaching motion.

Perception of hand motion direction uses a gravitational reference.

We studied possible frames of reference for kinesthetic perception of imposed hand motion direction in the frontal plane in ten young adult subjects with no history of neuromuscular disease. In one experiment, subjects were instructed to set unseen hand motion imposed by a motorized linear slide device parallel to the trunk-fixed longitudinal axis, seven visually specified axes and vertical (gravitational axis) while in a standard erect head/trunk posture and with head/trunk orientation varied. The visually specified axes were presented on a head-mounted display that also blocked vision of the external environment. In a second experiment using the same device, subjects set unseen hand motion parallel to vertical and to subjective oblique directions of 45 degrees clockwise (cw) and counter clockwise (ccw) from vertical in erect and varied head/trunk postures. Errors for setting hand motion to vertical and to verbally specified oblique axes (45 degrees cw and ccw from vertical) were lower than to the trunk longitudinal axis and visually specified axes. There were clear oblique effects in setting hand motion to visually specified axes and to subjective oblique (45 degrees cw and ccw) axes. When head and trunk orientation were varied, variable errors were higher for all axes, but remained lowest for vertical and subjective oblique axes. Moreover, errors for setting hand motion to all axes depended on head/trunk orientation. Overall, these results show that kinesthetic perception of imposed hand motion uses a subjective gravitational frame of reference that varies somewhat with head/trunk orientation.

Separating brain motion into rigid body displacement and deformation under low-severity impacts.

The relative motion of the brain with respect to the skull has been widely studied to investigate brain injury mechanisms under impacts, but the motion patterns are not yet thoroughly understood. This work analyzes brain motion patterns using the most recent and advanced experimental relative brain/skull motion data collected under low-severity impacts. With a minimum total pseudo-strain energy, the closed-form solutions for rigid body translation and rotation were obtained by matching measured neutral density target (NDT) positions with initial NDT positions. The brain motion was thus separated into rigid body displacement and deformation. The results show that the brain has nearly pure rigid body displacement at low impact speed. As the impact becomes more severe, the increased brain motion primarily is due to deformation, while the rigid body displacement is limited in magnitude for both translation and rotation. Under low-severity impacts in the sagittal plane, the rigid body brain translation has a magnitude of 4-5 mm, and the whole brain rotation is on the order of +/-5 degrees.