Level Ground Walking Research Articles

Patient-Preferred Prosthetic Ankle-Foot Alignment for Ramps and Level-Ground Walking

Patient preference of lower limb prosthesis behavior informally guides clinical decision making, and may become increasingly important for tuning new robotic prostheses. However, the processes for quantifying preference are still being developed, and the strengths and weaknesses of preference are not adequately understood. The present study sought to characterize the reliability (consistency) of patient preference of alignment during level-ground walking, and determine the patient-preferred ankle angle for ascent and descent of a 10° ramp, with implications for the design and control of robotic prostheses. Seven subjects with transtibial amputation walked over level ground, and ascended and descended a 10° ramp on a semi-active prosthetic ankle capable of unweighted repositioning in dorsiflexion and plantarflexion. Preferred ankle angle was measured with an adaptive forced-choice psychophysics paradigm, in which subjects walked on a randomized static ankle angle and reported whether they would prefer the ankle to be dorsiflexed or plantarflexed. Subjects had reliable preferences for alignment during level-ground walking, with deviations of 1.5° from preference resulting in an 84% response rate preferring changes toward the preference. Relative to level walking, subjects preferred 7.8° (SD: 4.8°) of dorsiflexion during ramp ascent, and 5.3° (SD: 3.8°) plantarflexion during ramp descent. As the ankle angle better matched the ramp angle, socket pressures and tibial progression (shank pitch) both more closely mirrored those during level walking. These findings provide baseline behaviors for prosthetic ankles capable of adapting to slopes based on patient preference, and provide strong evidence that people with transtibial amputation can finely perceive ankle alignment.

IMU-Based Deep Neural Networks: Prediction of Locomotor and Transition Intentions of an Osseointegrated Transfemoral Amputee.

This paper focuses on the design and comparison of different deep neural networks for the real-time prediction of locomotor and transition intentions of one osseointegrated transfemoral amputee using only data from inertial measurement units. The deep neural networks are based on convolutional neural networks, recurrent neural networks, and convolutional recurrent neural networks. The architectures' input are features in both the time domain and the time-frequency domain, which are derived from either one inertial measurement unit (placed above the prosthetic knee) or two inertial measurement units (placed above and below the prosthetic knee). The prediction of eight different locomotion modes (i.e., sitting, standing, level ground walking, stair ascent and descent, ramp ascent and descent, walking on uneven terrain) and the twenty-four transitions among them is investigated. The study shows that a recurrent neural network, realized with four layers of gated recurrent unit networks, achieves (with a 5-fold cross-validation) a mean F1 score of 84.78% and 86.50% using one inertial measurement unit, and 93.06% and 89.99% using two inertial measurement units, with or without sitting, respectively.

A Muscle Synergy-Inspired Method of Detecting Human Movement Intentions Based on Wearable Sensor Fusion.

Detecting human movement intentions is fundamental to neural control of robotic exoskeletons, as it is essential for achieving seamless transitions between different locomotion modes. In this study, we enhanced a muscle synergy-inspired method of locomotion mode identification by fusing the electromyography data with two types of data from wearable sensors (inertial measurement units), namely linear acceleration and angular velocity. From the finite state machine perspective, the enhanced method was used to systematically identify 2 static modes, 7 dynamic modes, and 27 transitions among them. In addition to the five broadly studied modes (level ground walking, ramps ascent/descent, stairs ascent/descent), we identified the transition between different walking speeds and modes of ramp walking at different inclination angles. Seven combinations of sensor fusion were conducted, on experimental data from 8 able-bodied adult subjects, and their classification accuracy and prediction time were compared. Prediction based on a fusion of electromyography and gyroscope (angular velocity) data predicted transitions earlier and with higher accuracy. All transitions and modes were identified with a total average classification accuracy of 94.5% with fused sensor data. For nearly all transitions, we were able to predict the next locomotion mode 300-500ms prior to the step into that mode.

Deep Reinforcement Learning for Physics-Based Musculoskeletal Simulations of Healthy Subjects and Transfemoral Prostheses' Users During Normal Walking.

This paper proposes to use deep reinforcement learning for the simulation of physics-based musculoskeletal models of both healthy subjects and transfemoral prostheses' users during normal level-ground walking. The deep reinforcement learning algorithm is based on the proximal policy optimization approach in combination with imitation learning to guarantee a natural walking gait while reducing the computational time of the training. Firstly, the optimization algorithm is implemented for the OpenSim model of a healthy subject and validated with experimental data from a public data-set. Afterwards, the optimization algorithm is implemented for the OpenSim model of a generic transfemoral prosthesis' user, which has been obtained by reducing the number of muscles around the knee and ankle joints and, specifically, by keeping only the uniarticular ones. The model of the transfemoral prosthesis' user shows a stable gait, with a forward dynamic comparable to the healthy subject's, yet using higher muscles' forces. Even though the computed muscles' forces could not be directly used as control inputs for muscle-like linear actuators due to their pattern, this study paves the way for using deep reinforcement learning for the design of the control architecture of transfemoral prostheses.

A Locomotion Mode Adaptive Strategy for Real-Time Detection of Gait Events During Negotiating Staircases

The joint kinematics for negotiating different locomotion modes (LMs), mainly level-ground walking (LGW) and ascending and descending staircases, are significantly different. Therefore, an algorithm that could detect gait events during the LGW would not be efficient during the stair ascent (SA) and stair descent (SD) even though the one designed for SA would not be suitable for SD, and vice versa. In this study, a novel gait event detection (GED) strategy is proposed, which is adaptive to the change in LM. This adaptability provided the flexibility of using different sets of rules for different LMs. For this purpose, an LM classification strategy was designed that supervised the selection of the rule-base for three LMs. The rules were applied to the kinematic data collected using two inertial sensors attached to the shank and foot of the subject. This method was validated with ten control subjects recruited for the experiment. The proposed method was seen to improve the overall performance of GED compared with the existing methods that used the same algorithm for all three LMs.

Prosthetic Leg Locomotion-Mode Identification Based on High-Order Zero-Crossing Analysis Surface Electromyography

The research purpose was to improve the accuracy in identifying the prosthetic leg locomotion mode. Surface electromyography (sEMG) combined with high-order zero-crossing was used to identify the prosthetic leg locomotion modes. sEMG signals recorded from residual thigh muscles were chosen as inputs to pattern classifier for locomotion-mode identification. High-order zero-crossing were computed as the sEMG features regarding locomotion modes. Relevance vector machine (RVM) classifier was investigated. Bat algorithm (BA) was used to compute the RVM classifier kernel function parameters. The classification performance of the particle swarm optimization-relevance vector machine (PSO-RVM) and RVM classifiers was compared. The BA-RVM produced lower classification error in sEMG pattern recognition for the transtibial amputees over a variety of locomotion modes: upslope, downgrade, level-ground walking and stair ascent/descent.

Machine Learning Approaches for Activity Recognition and/or Activity Prediction in Locomotion Assistive Devices-A Systematic Review.

Locomotion assistive devices equipped with a microprocessor can potentially automatically adapt their behavior when the user is transitioning from one locomotion mode to another. Many developments in the field have come from machine learning driven controllers on locomotion assistive devices that recognize/predict the current locomotion mode or the upcoming one. This review synthesizes the machine learning algorithms designed to recognize or to predict a locomotion mode in order to automatically adapt the behavior of a locomotion assistive device. A systematic review was conducted on the Web of Science and MEDLINE databases (as well as in the retrieved papers) to identify articles published between 1 January 2000 to 31 July 2020. This systematic review is reported in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines and is registered on Prospero (CRD42020149352). Study characteristics, sensors and algorithms used, accuracy and robustness were also summarized. In total, 1343 records were identified and 58 studies were included in this review. The experimental condition which was most often investigated was level ground walking along with stair and ramp ascent/descent activities. The machine learning algorithms implemented in the included studies reached global mean accuracies of around 90%. However, the robustness of those algorithms seems to be more broadly evaluated, notably, in everyday life. We also propose some guidelines for homogenizing future reports.

On-board Training Strategy for IMU-Based Real-Time Locomotion Recognition of Transtibial Amputees With Robotic Prostheses.

The paper puts forward an on-board strategy for a training model and develops a real-time human locomotion mode recognition study based on a trained model utilizing two inertial measurement units (IMUs) of robotic transtibial prosthesis. Three transtibial amputees were recruited as subjects in this study to finish five locomotion modes (level ground walking, stair ascending, stair descending, ramp ascending, and ramp descending) with robotic prostheses. An interaction interface was designed to collect sensors' data and instruct to train model and recognition. In this study, the analysis of variance ratio (no more than 0.05) reflects the good repeatability of gait. The on-board training time for SVM (Support Vector Machines), QDA (Quadratic Discriminant Analysis), and LDA (Linear discriminant analysis) are 89, 25, and 10 s based on a 10,000 × 80 training data set, respectively. It costs about 13.4, 5.36, and 0.067 ms for SVM, QDA, and LDA for each recognition process. Taking the recognition accuracy of some previous studies and time consumption into consideration, we choose QDA for real-time recognition study. The real-time recognition accuracies are 97.19 ± 0.36% based on QDA, and we can achieve more than 95% recognition accuracy for each locomotion mode. The receiver operating characteristic also shows the good quality of QDA classifiers. This study provides a preliminary interaction design for human–machine prosthetics in future clinical application. This study just adopts two IMUs not multi-type sensors fusion to improve the integration and wearing convenience, and it maintains comparable recognition accuracy with multi-type sensors fusion at the same time.

Biological Hip Torque Estimation using a Robotic Hip Exoskeleton.

Machine learning (ML) algorithms present an opportunity to estimate joint kinetics using a limited set of mechanical sensors. These estimates could be used as a continuous reference signal for exoskeleton control, able to modulate exoskeleton assistance in real-world environments. In this study, sagittal plane biological hip torque during level ground, incline and decline walking was calculated using inverse dynamics of human subject data. Subsequently, this torque was estimated using neural network (NN) and XGBoost ML models. Model inputs consisted solely of mechanical sensor data onboard a robotic hip exoskeleton. These results were compared to a baseline method of estimating hip torque as the mean torque profile during ambulation. On average across conditions, the NN and XGBoost models estimated biological hip torque with an RMSE of 0.116±0.015 and 0.108±0.011 Nm/kg, respectively, which was significantly less than the baseline estimation that had an RMSE of 0.300±0.145 Nm/kg (p<0.05). Fitting the baseline method to ambulation mode specific data significantly reduced overall RMSE by 59.3%; however, the ML models were still significantly better than the baseline method (p<0.05). These results show that machine learning algorithms can estimate biological hip torque using only mechanical sensors onboard a hip exoskeleton better than simply using an average torque profile. This suggests that these estimation models could be suitable for modulating exoskeleton assistance. Additionally, no evidence suggested the need to train separate ML models for each ambulation mode as estimation RMSE was not significantly different across unified and separated ML models.

Flexible Dry Electrodes for EMG Acquisition within Lower Extremity Prosthetic Sockets.

Acquisition of surface electromyography (sEMG) from a person with an amputated lower extremity (LE) during prosthesis-assisted walking remains a significant challenge due to the dynamic nature of the gait cycle. Current solutions to sEMG-based neural control of active LE prostheses involve a combination of customized electrodes, prosthetic sockets, and liners. These technologies are generally: (i) incompatible with a subject's existing prosthetic socket and liners; (ii) uncomfortable to use; and (iii) expensive. This paper presents a flexible dry electrode design for sEMG acquisition within LE prosthetic sockets which seeks to address these issues. Design criteria and corresponding design decisions are explained and a proposed flexible electrode prototype is presented. Performances of the proposed electrode and commercial Ag/AgCl electrodes are compared in seated subjects without amputations. Quantitative analyses suggest comparable signal qualities for the proposed novel electrode and commercial electrodes. The proposed electrode is demonstrated in a subject with a unilateral transtibial amputation wearing her own liner, socket, and the portable sEMG processing platform in a preliminary standing and level ground walking study. Qualitative analyses suggest the feasibility of real-time sEMG data collection from load-bearing, ambulatory subjects.

Assessment of Intuitiveness and Comfort of Wearable Haptic Feedback Strategies for Assisting Level and Stair Walking

Nowadays, lower-limb prostheses are reaching real-world usability especially on ground-level walking. However, some key tasks such as stair walking are still quite demanding. Providing haptic feedback about the foot placement on the steps might reduce the cognitive load of the task, compensating for increased dependency on vision and lessen the risk of falling. Experiments on intact subjects can be useful to define the feedback strategies prior to clinical trials, but effective methods to assess the efficacy of the strategies are few and usually rely on the emulation of the disability condition. The present study reports on the design and testing of a wearable haptic feedback system in a protocol involving intact subjects to assess candidate strategies to be adopted in clinical trials. The system integrated a sensorized insole wirelessly connected to a textile waist belt equipped with three vibrating motors. Three stimulation strategies for mapping the insole pressure data to vibrotactile feedback were implemented and compared in terms of intuitiveness and comfort perceived during level and stair walking. The strategies were ranked using a relative rating approach, which highlighted the differences between them and suggested guidelines for their improvement. The feedback evaluation procedure proposed could facilitate the selection and improvement of haptic feedback strategies prior to clinical testing.

An Infrared Sensor-Based Instrumented Shoe for Gait Events Detection on Different Terrains and Transitions

This paper presents a low-cost wireless gait event detection system that incorporates infra-red range sensors mounted laterally at a specific anatomical location on the shoe. This system uses foot clearance information to extract heel strike (HS) and toe-off (TO) events. A subject-specific algorithm based on signature curve matching was developed to estimate the HS and TO in six able-bodied subjects and three transfemoral amputees. The data were acquired in a real-life environment while the subject walked on level ground and ramp terrains, including the transitions. The HS and TO in the level ground were detected with an accuracy of 22.87± 9.93 ms and 10.42± 6.70 ms, respectively, in able-bodied participants. The accuracy of 37.91± 12.36 ms and 20.41± 11.23 ms in HS and TO, respectively, was observed in amputees during level ground walking. Further, within the ramp terrain, the HS and TO were detected with an accuracy of 23.33± 11.20 ms and 13.36± 9.5 ms respectively in able-bodied subjects and 40.35± 12.64 ms and 47.51± 19.23 ms in amputees. This method reports comparable accuracy, with minimal standard deviation, to other existing methods based on accelerometer, gyroscope, electromyography, and force myography. This system is robust, easy to operate, and avoids direct contact with the body. The developed system offers a movable sensor platform in two degrees of freedom that facilitates the adaptability of the proposed system for different heel height and shoe sizes.

Walking Strategies and Performance Evaluation for Human-Exoskeleton Systems under Admittance Control.

Lower-limb exoskeletons as walking assistive devices have been intensively investigated in recent decades. In these studies, intention detection and performance evaluation are important topics. In our previous studies, we proposed a disturbance observer (DOB)-based torque estimation algorithm and an admittance control law to shape the admittance of the human-exoskeleton system (HES) and comply with the user’s walking intention. These algorithms have been experimentally verified under the condition of no ground reaction force (GRF) in our previous studies. In this paper, we devised and integrated with the exoskeleton control system a sensing and communication module on each foot to measure and compensate for GRF. Rigorous theoretical analysis was performed and the sufficient conditions for the robust stability of the closed-loop system were derived. Then, we conducted level ground assistive walking repeatedly with different test subjects and exhaustive combinations of admittance parameters. In addition, we proposed two tractable and physically insightful performance indices called normalized energy consumption index (NECI) and walking distance in a fixed period of time to quantitatively evaluate the performance for different admittance parameters. We also compared the energy consumption for users walking with and without the exoskeleton. The results show that the proposed admittance control law reduces the energy consumption of the user during level ground walking.

Continuous Prediction of Joint Angular Positions and Moments: A Potential Control Strategy for Active Knee-Ankle Prostheses

Transfemoral amputation substantially impairs locomotion. To restore the lost locomotive capability, amputees rely on knee-ankle prostheses. Theoretically, active knee-ankle prostheses may better support natural gait than their passive counterparts by replacing the missing muscle function. The control algorithms of such active devices need to comprehend the user's locomotive intention and convert them into control commands for actuating the prosthesis. For an active knee-ankle prosthesis, the gait variables to be controlled to allow the desired locomotion could be the knee angle, knee moment, ankle angle, ankle moment. In this paper, a random forest regression model is employed for the continuous prediction of these gait variables for level ground walking at self-selected normal speed. Experimentally obtained thigh kinematics were the input to the random forest model. The proposed method could predict the angles and moments of the knee and ankle with high accuracy (mean $R^{2}$ value of 0.97 for ankle angle, 0.98 for ankle moment, 0.99 for knee angle, 0.95 for knee moment across four able-bodied subjects). The proposed strategy shows potential for continuously controlling an active knee-ankle prosthesis for transfemoral amputees, whose thigh angular motion can be used to infer the required prosthetic moments or angles.

A Backpack Minimizing the Vertical Acceleration of the Load Improves the Economy of Human Walking.

Loaded walking with a rucksack results in both gravitational and inertial forces of the load that must be borne by human carriers. The inertial force may be the source of metabolic burden and musculoskeletal injuries. This paper presents a lightweight backpack with a disturbance observer-based acceleration control to minimize the inertial force. The backpack was evaluated by seven participants walking on a treadmill at 5 km h-1 with a 19.4 kg load. Three experimental conditions were involved, including walking with a locked load (LOCKED), with an acceleration-controlled load (ACTIVE) using the designed backpack and walking with the same load using a rucksack (RUCKSACK). Our results showed that the ACTIVE condition reduces the load acceleration by 98.5% on average, and reduce the gross metabolic power by 8.0% and 11.0% as compared to LOCKED and RUCKSACK conditions respectively. The results demonstrate that the proposed active backpack can improve the loaded walking economy compared with a conventional rucksack in level-ground walking.

A biomechanical analysis of 3D stress and strain patterns in patellar tendon during knee flexion.

Patellar tendinopathy is among the most widespread patellar tendon diseases in athletes that participate in activities involving running and jumping. Although their symptoms can be detected, especially at the inferior pole of the patella, their biomechanical cause remains unknown. In this study, a three-dimensional finite element model of knee complex was developed to investigate principal stress and strain distributions in the patellar tendon during 0° to 90° knee flexion and slow and fast level-ground walking. Results indicate that the patellar tendon is subjected to tensile stress and strains during all three activities. During flexion, its central proximal posterior region exhibited highest peak stress and strain, followed by central distal posterior, central distal anterior and central proximal anterior region. Similar trends and magnitudes were reported during slow and fast walking. The region with highest principal stresses and strains, central proximal anterior region, also corresponds to the most commonly reported patellar tendinopathy lesion site, suggesting that principal stress and strain are good indicators of lesion site location. Additional factors such as regional variations in material properties and frequency and duration of cyclic loading also need to be considered when determining the biomechanical aetiology of patellar tendinopathy.

Real-Time Detection of Actual and Early Gait Events During Level-Ground and Ramp Walking

In recent studies on gait event detection, the inertial sensor-based approaches were found to be more suitable for long term use compared to force sensors. However, inertial sensor-based algorithms exhibit some time delay. This time delay can affect the stability of the powered prosthetic and assistive devices (PPADs). Hence, several methods were proposed to reduce the same. However, it is impossible to completely eliminate the effect of delay using the existing methods. In this study, we proposed a novel early detection strategy to tackle the issue discussed above. The strategy involved the detection of both actual as well as early gait events. Thus, a total of nine gait events were detected in real-time using inertial sensors and a set of rules. The proposed method was tested for ten control subjects during both level ground as well as ramp walking, and the detection accuracy was found to be equal to 100%. Moreover, the time delay for the detection of actual gait events using the proposed method was found to be similar to that of some state-of-the-art methods and better than a few. The average early detection time was found to be equal to 123 ms that can be utilized to improve the response of PPADs.

Powered knee and ankle prosthesis with indirect volitional swing control enables level-ground walking and crossing over obstacles.

Powered prostheses aim to mimic the missing biological limb with controllers that are finely tuned to replicate the nominal gait pattern of non-amputee individuals. Unfortunately, this control approach poses a problem with real-world ambulation, which includes tasks such as crossing over obstacles, where the prosthesis trajectory must be modified to provide adequate foot clearance and ensure timely foot placement. Here, we show an indirect volitional control approach that enables prosthesis users to walk at different speeds while smoothly and continuously crossing over obstacles of different sizes without explicit classification of the environment. At the high level, the proposed controller relies on a heuristic algorithm to continuously change the maximum knee flexion angle and the swing duration in harmony with the user's residual limb. At the low level, minimum-jerk planning is used to continuously adapt the swing trajectory while maximizing smoothness. Experiments with three individuals with above-knee amputation show that the proposed control approach allows for volitional control of foot clearance, which is necessary to negotiate environmental barriers. Our study suggests that a powered prosthesis controller with intrinsic, volitional adaptability may provide prosthesis users with functionality that is not currently available, facilitating real-world ambulation.

Image Transformation and CNNs: A Strategy for Encoding Human Locomotor Intent for Autonomous Wearable Robots

Wearable robots have the potential to improve the lives of countless individuals; however, challenges associated with controlling these systems must be addressed before they can reach their full potential. Modern control strategies for wearable robots are predicated on activity-specific implementations, and testing is usually limited to a single, fixed activity within the laboratory (e.g., level ground walking). To accommodate various activities in real-world scenarios, control strategies must include the ability to safely and seamlessly transition between activity-specific controllers. One potential solution to this challenge is to the infer wearer's intent using pattern recognition of locomotion sensor data. To this end, we developed an intent recognition framework implementing convolutional neural networks with image encoding (i.e. spectrogram) that enables prediction of the upcoming locomotor activity of the wearer's next step. In this letter, we describe our intent recognition system, comprised of a mel-spectrogram and subsequent neural network architecture. In addition, we analyzed the effect of sensor locations and modalities on the recognition system, and compared our proposed system to state-of-the-art locomotor intent recognition strategies. We were able to attain high classification performance (error rate: 1.1%), which was comparable or better than previous systems.

Loads at the Implant-Prosthesis Interface During Free and Aided Ambulation in Osseointegrated Transfemoral Prostheses

Bone-anchored attachment of amputation limb prostheses is increasingly becoming a clinically accepted alternative to conventional socket suspension. The direct transfer of loads demands that the percutaneous implant system and the residual bone withstand all forces and moments transferred from the prosthesis. This study presents load measurements recorded at the bone-anchored attachment in 20 individuals with unilateral transfemoral amputation performing the everyday ambulatory activities: level ground walking, stairs ascent/descent and slope ascent/descent. Mean peak values for the sample populations across activities ranged from 498-684 N for the resultant force, 26.5-39.8 Nm for the bending moment, and 3.1-5.5 Nm for the longitudinal moment. Significant differences with respect to level walking were found for the resultant force during stairs ascent, (higher, p = 0.002), and stairs descent, (lower, p = 0.005). Using a crutch reduced the peak resultant forces and the peak bending moments with averages ranging from 5.5-12.6 % and 13.2-15.6 %, respectively. Large inter-participant variations were observed and no single activity resulted in consistently higher loading of the bone-anchored attachment across the participants. Results from this study can guide future development of percutaneous osseointegrated implant systems for limb prostheses and their rehabilitation protocols.