Peripheral Arterial Disease
Conditions
Keywords
Exoskeleton
Brief summary
Exoskeletons, wearable devices that assist with walking, can improve mobility in clinical populations. With exoskeletons, it is crucial to optimize the assistance profile. Recent studies describe algorithms (i.e., human-in-the-loop) to optimize the assistance profile with real-time metabolic measurements. The needed duration of current human-in-the-loop (HITL) algorithms range from 20 minutes to 1 hour which is longer than the average duration that most patients with peripheral artery disease (PAD) can walk. Because of this limited walking duration, it is often not possible for patients with PAD to reach steady-state metabolic cost, which makes these measurements are not useful for optimizing exoskeletons. In this study, investigators intend to develop and evaluate HITL optimization methods for exoskeletons and use the information to design and evaluate a portable hip exoskeleton. Shorter and more clinically feasible HITL optimization strategies based on experiments in healthy adults might allow utilizing these optimization strategies to become available for patient populations such as patients with PAD.
Detailed description
Exoskeletons, wearable devices that assist with walking, can improve mobility in clinical populations. With exoskeletons, it is crucial to optimize the assistance profile. Recent studies describe algorithms (i.e., human-in-the-loop) to optimize the assistance profile with real-time metabolic measurements. The needed duration of current human-in-the-loop (HITL) algorithms range from 20 minutes to 1 hour which is longer than the average duration that most patients with peripheral artery disease (PAD) can walk. Because of this limited walking duration, it is often not possible for patients with PAD to reach steady-state metabolic cost, which makes these measurements are not useful for optimizing exoskeletons. Shorter and more clinically feasible HITL optimization strategies based on experiments in healthy adults might allow utilizing these optimization strategies to become available for patient populations such as patients with PAD. This study will test different methods for optimizing exoskeletons. It will consist of an habituation session to the hip exoskeleton, an optimization session to find the optimal actuation settings using an algorithm that converges toward the optimum based on real-time measurements (human-in-the-loop algorithm) and a post-test at the end of optimization session to compare different conditions. The outcomes will be evaluated by surface electromyography, exoskeleton sensors, ground reaction force, walking speed, indirect calorimetry, and motion capture (Vicon).
Interventions
OTHERExoskeleton Optimization
Participants will walk 10-minute trials while an optimization algorithm changes the assistance profile of the exoskeleton.
OTHEREndurance Evaluation
Participants will walk 2 trials at a speed of 1 meter per second until the participant indicates claudication or a maximum duration of 6 minutes, which ever comes first.
Sponsors
National Institute of General Medical Sciences (NIGMS)
University of Nebraska
Study design
Allocation
NON_RANDOMIZED
Intervention model
PARALLEL
Primary purpose
BASIC_SCIENCE
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
19 Years to 85 Years
Healthy volunteers
Yes
Inclusion criteria
* Ability to provide written consent * Chronic claudication history * Ankle-brachial index \< 0.90 at rest * Stable blood pressure, lipids, and diabetes for \> 6 weeks * Ability to walk on a treadmill for multiple five-minute spans * Ability to fit in exoskeleton * Waist circumference 78 to 92 centimeters (31 to 36 inches) * Thigh circumference 48 to 60 centimeters (19 to 24 inches) * Minimal thigh length 28 centimeters (11 inches)
Exclusion criteria
* Resting pain or tissue loss due to peripheral artery disease (PAD, Fontaine stage III and IV) * Foot ulceration * Acute lower extremity event secondary to thromboembolic disease or acute trauma * Walking capacity limited by diseases unrelated to PAD, such as: * Neurological disorders * Musculoskeletal disorders (arthritis, scoliosis, stroke, spinal injury, etc.) * History of ankle instability * Knee injury * Diagnosed joint laxity * Lower limb injury * Surgery within the past 12 months * Joint replacement * Pulmonary disease or breathing disorders * Cardiovascular disease * Vestibular disorder * Acute injury or pain in lower extremity * Current illness * Inability to follow visual cues due to blindness * Inability to follow auditory cues due to deafness * Pregnant
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Time to Convergence | 10 minutes | Convergence is determined when the estimated optimal exoskeleton settings vary less than 10%. The time to convergence is measured. |
| Peak Extension Timing | 20 seconds | The time to peak extension moment of exoskeleton is measured by plotting the exoskeleton moment versus stride cycle percentage and finding the timing when the peak in the extension moment occurs expressed in percent of the stride cycle. |
| Peak Flexion Timing | 20 seconds | The time to peak flexion moment of exoskeleton is measured by plotting the flexion moment versus stride cycle percentage and finding the timing when the peak in the flexion moment occurs expressed in percent of the stride cycle. |
| Largest Lyapunov Exponent | 20 seconds | Largest Lyapunov exponent (the rate of separation of infinitesimally close trajectories) of lower limb kinematics is determined. Largest Lyapunov exponent is calculated using Wolf's algorithm. The theoretical range is from zero to plus infinity. Zero indicates an entirely stable periodic movement pattern. Higher values indicate more unstable and chaotic movement patterns. Lower values are considered better, and higher values are considered worse for gait stability. |
Countries
United States
Participant flow
Pre-assignment details
No enrolled participants were tested in the endurance arm because the preceding optimal assistance arm was not successful based on the convergence criteria.
Participants by arm
| Arm | Count |
|---|---|
| Optimal Assistance Pattern An optimization algorithm will change the assistance pattern on the hip exoskeleton during walking sessions and the optimal assistance pattern will be determined when gait variability is minimized.
Exoskeleton Optimization: Participants will walk 10-minute trials while an optimization algorithm changes the assistance profile of an exoskeleton. | 9 |
| Effects on Endurance The effects on endurance of participants using ground reaction force (Bertec treadmill), walking speed (Bertec treadmill), indirect calorimetry (Cosmed), and motion capture (Vicon) will be determined.
Endurance Evaluation: Participants will walk 2 trials at a speed of 1 meter per second until the participant indicates claudication or a maximum duration of 6 minutes. | 0 |
| Total | 9 |
Baseline characteristics
| Characteristic | Optimal Assistance Pattern | Effects on Endurance | Total |
|---|---|---|---|
| Age, Categorical <=18 years | 0 Participants | 0 Participants | 0 Participants |
| Age, Categorical >=65 years | 1 Participants | 0 Participants | 1 Participants |
| Age, Categorical Between 18 and 65 years | 8 Participants | 0 Participants | 8 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Asian | 2 Participants | 0 Participants | 2 Participants |
| Race (NIH/OMB) Black or African American | 1 Participants | 0 Participants | 1 Participants |
| Race (NIH/OMB) More than one race | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 1 Participants | 0 Participants | 1 Participants |
| Race (NIH/OMB) White | 5 Participants | 0 Participants | 5 Participants |
| Region of Enrollment United States | 9 participants | — | 9 participants |
| Sex: Female, Male Female | 4 Participants | 0 Participants | 4 Participants |
| Sex: Female, Male Male | 5 Participants | 0 Participants | 5 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 9 | 0 / 0 |
| other Total, other adverse events | 0 / 9 | 0 / 0 |
| serious Total, serious adverse events | 0 / 9 | 0 / 0 |
Outcome results
Primary
Largest Lyapunov Exponent
Largest Lyapunov exponent (the rate of separation of infinitesimally close trajectories) of lower limb kinematics is determined. Largest Lyapunov exponent is calculated using Wolf's algorithm. The theoretical range is from zero to plus infinity. Zero indicates an entirely stable periodic movement pattern. Higher values indicate more unstable and chaotic movement patterns. Lower values are considered better, and higher values are considered worse for gait stability.
Time frame: 20 seconds
Population: The effect on endurance arm was not analyzed since the preceding optimal assistance pattern aim was not successful based on the predefined convergence criteria.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Optimal Assistance Pattern | Largest Lyapunov Exponent | 6.6 (Lyapunov exponent is unitless) | Standard Deviation 2.8 |
Primary
Peak Extension Timing
The time to peak extension moment of exoskeleton is measured by plotting the exoskeleton moment versus stride cycle percentage and finding the timing when the peak in the extension moment occurs expressed in percent of the stride cycle.
Time frame: 20 seconds
Population: The effect on endurance arm was not analyzed since the preceding optimal assistance pattern aim was not successful based on the predefined convergence criteria.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Optimal Assistance Pattern | Peak Extension Timing | 86 % stride cycle | Standard Deviation 8 |
Primary
Peak Flexion Timing
The time to peak flexion moment of exoskeleton is measured by plotting the flexion moment versus stride cycle percentage and finding the timing when the peak in the flexion moment occurs expressed in percent of the stride cycle.
Time frame: 20 seconds
Population: The effect on endurance arm was not analyzed since the preceding optimal assistance pattern aim was not successful based on the predefined convergence criteria.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Optimal Assistance Pattern | Peak Flexion Timing | 56 % stride cycle | Standard Deviation 2 |
Primary
Time to Convergence
Convergence is determined when the estimated optimal exoskeleton settings vary less than 10%. The time to convergence is measured.
Time frame: 10 minutes
Population: The time to convergence for the optimal assistance pattern could not be reported since the convergence criterion was not achieved. The effect on endurance arm was not analyzed since the preceding optimal assistance pattern aim was not successful based on the predefined convergence criteria.