Rapid energy expenditure estimation for ankle assisted and inclined loaded walking

Human-in-the-loop optimization of wearable device parameters using an EMG-based objective function

Advancements in wearable robots aim to improve user motion, motor control, and overall experience by minimizing energetic cost (EC). However, EC is challenging to measure and it is typically indirectly estimated through respiratory gas analysis. This study introduces a novel EMG-based objective function that captures individuals' natural energetic expenditure during walking. The objective function combines information from electromyography (EMG) variables such as intensity and muscle synergies. First, we demonstrate the similarity of the proposed objective function, calculated offline, to the EC during walking. Second, we minimize and validate the EMG-based objective function using an online Bayesian optimization algorithm. The walking step frequency is chosen as the parameter to optimize in both offline and online approaches in order to simplify experiments and facilitate comparisons with related research. Compared to existing studies that use EC as the objective function, results demonstrated that the optimization of the presented objective function reduced the number of iterations and, when compared with gradient descent optimization strategies, also reduced convergence time. Moreover, the algorithm effectively converges toward an optimal step frequency near the user's preferred frequency, positively influencing EC reduction. The good correlation between the estimated objective function and measured EC highlights its consistency and reliability. Thus, the proposed objective function could potentially optimize lower limb exoskeleton assistance and improve user performance and human-robot interaction without the need for challenging respiratory gas measurements.

Estimating human joint moments unifies exoskeleton control, reducing user effort

Robotic lower-limb exoskeletons can augment human mobility, but current systems require extensive, context-specific considerations, limiting their real-world viability. Here, we present a unified exoskeleton control framework that autonomously adapts assistance on the basis of instantaneous user joint moment estimates from a temporal convolutional network (TCN). When deployed on our hip exoskeleton, the TCN achieved an average root mean square error of 0.142 newton-meters per kilogram across 35 ambulatory conditions without any user-specific calibration. Further, the unified controller significantly reduced user metabolic cost and lower-limb positive work during level-ground and incline walking compared with walking without wearing the exoskeleton. This advancement bridges the gap between in-lab exoskeleton technology and real-world human ambulation, making exoskeleton control technology viable for a broad community.

IMU-Based Energy Expenditure Estimation for Various Walking Conditions Using a Hybrid CNN-LSTM Model

In ubiquitous healthcare systems, energy expenditure estimation based on wearable sensors such as inertial measurement units (IMUs) is important for monitoring the intensity of physical activity. Although several studies have reported data-driven methods to estimate energy expenditure during activities of daily living using wearable sensor signals, few have evaluated the performance while walking at various speeds and inclines. In this study, we present a hybrid model comprising a convolutional neural network (CNN) and long short-term memory (LSTM) to estimate the steady-state energy expenditure under various walking conditions based solely on IMU data. To implement and evaluate the model, we performed level/inclined walking and level running experiments on a treadmill. With regard to the model inputs, the performance of the proposed model based on fixed-size sequential data was compared with that of a method based on stride-segmented data under different conditions in terms of the sensor location, input sequence format, and neural network model. Based on the experimental results, the following conclusions were drawn: (i) the CNN-LSTM model using a two-second sequence from the IMU attached to the lower body yielded optimal performance, and (ii) although the stride-segmented data-based method showed superior performance, the performance difference between the two methods was not significant; therefore, the proposed model based on fixed-size sequential data may be considered more practical as it does not require heel-strike detection.

Opportunities and challenges in the development of exoskeletons for locomotor assistance

Exoskeletons can augment the performance of unimpaired users and restore movement in individuals with gait impairments. Knowledge of how users interact with wearable devices and of the physiology of locomotion have informed the design of rigid and soft exoskeletons that can specifically target a single joint or a single activity. In this Review, we highlight the main advances of the past two decades in exoskeleton technology and in the development of lower-extremity exoskeletons for locomotor assistance, discuss research needs for such wearable robots and the clinical requirements for exoskeleton-assisted gait rehabilitation, and outline the main clinical challenges and opportunities for exoskeleton technology.

Muscle coordination and recruitment during squat assistance using a robotic ankle-foot exoskeleton

Squatting is an intensive activity routinely performed in the workplace to lift and lower loads. The effort to perform a squat can decrease using an exoskeleton that considers individual worker's differences and assists them with a customized solution, namely, personalized assistance. Designing such an exoskeleton could be improved by understanding how the user's muscle activity changes when assistance is provided. This study investigated the change in the muscle recruitment and activation pattern when personalized assistance was provided. The personalized assistance was provided by an ankle-foot exoskeleton during squatting and we compared its effect with that of the no-device and unpowered exoskeleton conditions using previously collected data. We identified four main muscle recruitment strategies across ten participants. One of the strategies mainly used quadriceps muscles, and the activation level corresponding to the strategy was reduced under exoskeleton assistance compared to the no-device and unpowered conditions. These quadriceps dominant synergy and rectus femoris activations showed reasonable correlations (r = 0.65, 0.59) to the metabolic cost of squatting. These results indicate that the assistance helped reduce quadriceps activation, and thus, the metabolic cost of squatting. These outcomes suggest that the muscle recruitment and activation patterns could be used to design an exoskeleton and training methods.

Ambient Assisted Living: Scoping Review of Artificial Intelligence Models, Domains, Technology, and Concerns

BACKGROUND

Ambient assisted living (AAL) is a common name for various artificial intelligence (AI)-infused applications and platforms that support their users in need in multiple activities, from health to daily living. These systems use different approaches to learn about their users and make automated decisions, known as AI models, for personalizing their services and increasing outcomes. Given the numerous systems developed and deployed for people with different needs, health conditions, and dispositions toward the technology, it is critical to obtain clear and comprehensive insights concerning AI models used, along with their domains, technology, and concerns, to identify promising directions for future work.

OBJECTIVE

This study aimed to provide a scoping review of the literature on AI models in AAL. In particular, we analyzed specific AI models used in AАL systems, the target domains of the models, the technology using the models, and the major concerns from the end-user perspective. Our goal was to consolidate research on this topic and inform end users, health care professionals and providers, researchers, and practitioners in developing, deploying, and evaluating future intelligent AAL systems.

METHODS

This study was conducted as a scoping review to identify, analyze, and extract the relevant literature. It used a natural language processing toolkit to retrieve the article corpus for an efficient and comprehensive automated literature search. Relevant articles were then extracted from the corpus and analyzed manually. This review included 5 digital libraries: IEEE, PubMed, Springer, Elsevier, and MDPI.

RESULTS

We included a total of 108 articles. The annual distribution of relevant articles showed a growing trend for all categories from January 2010 to July 2022. The AI models mainly used unsupervised and semisupervised approaches. The leading models are deep learning, natural language processing, instance-based learning, and clustering. Activity assistance and recognition were the most common target domains of the models. Ambient sensing, mobile technology, and robotic devices mainly implemented the models. Older adults were the primary beneficiaries, followed by patients and frail persons of various ages. Availability was a top beneficiary concern.

CONCLUSIONS

This study presents the analytical evidence of AI models in AAL and their domains, technologies, beneficiaries, and concerns. Future research on intelligent AAL should involve health care professionals and caregivers as designers and users, comply with health-related regulations, improve transparency and privacy, integrate with health care technological infrastructure, explain their decisions to the users, and establish evaluation metrics and design guidelines.

TRIAL REGISTRATION

PROSPERO (International Prospective Register of Systematic Reviews) CRD42022347590; https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022347590.

Deep Learning-Based Energy Expenditure Estimation in Assisted and Non-Assisted Gait Using Inertial, EMG, and Heart Rate Wearable Sensors

Energy expenditure is a key rehabilitation outcome and is starting to be used in robotics-based rehabilitation through human-in-the-loop control to tailor robot assistance towards reducing patients’ energy effort. However, it is usually assessed by indirect calorimetry which entails a certain degree of invasiveness and provides delayed data, which is not suitable for controlling robotic devices. This work proposes a deep learning-based tool for steady-state energy expenditure estimation based on more ergonomic sensors than indirect calorimetry. The study innovates by estimating the energy expenditure in assisted and non-assisted conditions and in slow gait speeds similarly to impaired subjects. This work explores and benchmarks the long short-term memory (LSTM) and convolutional neural network (CNN) as deep learning regressors. As inputs, we fused inertial data, electromyography, and heart rate signals measured by on-body sensors from eight healthy volunteers walking with and without assistance from an ankle-foot exoskeleton at 0.22, 0.33, and 0.44 m/s. LSTM and CNN were compared against indirect calorimetry using a leave-one-subject-out cross-validation technique. Results showed the suitability of this tool, especially CNN, that demonstrated root-mean-squared errors of 0.36 W/kg and high correlation (ρ > 0.85) between target and estimation (R¯2 = 0.79). CNN was able to discriminate the energy expenditure between assisted and non-assisted gait, basal, and walking energy expenditure, throughout three slow gait speeds. CNN regressor driven by kinematic and physiological data was shown to be a more ergonomic technique for estimating the energy expenditure, contributing to the clinical assessment in slow and robotic-assisted gait and future research concerning human-in-the-loop control.

A simple method reveals minimum time required to quantify steady-rate metabolism and net cost of transport for human walking

The U-shaped net cost of transport (COT) curve of walking has helped scientists understand the biomechanical basis that underlies energy minimization during walking. However, to produce an individual's net COT curve, data must be analyzed during periods of steady-rate metabolism. Traditionally, studies analyze the last few minutes of a 6-10 min trial, assuming that steady-rate metabolism has been achieved. Yet, it is possible that an individual achieves steady rates of metabolism much earlier. However, there is no consensus on how to objectively quantify steady-rate metabolism across a range of walking speeds. Therefore, we developed a simple slope method to determine the minimum time needed for humans to achieve steady rates of metabolism across slow to fast walking speeds. We hypothesized that a shorter time window could be used to produce a net COT curve that is comparable to the net COT curve created using traditional methods. We analyzed metabolic data from twenty-one subjects who completed several 7-min walking trials ranging from 0.50-2.00 m/s. We partitioned the metabolic data for each trial into moving 1-min, 2-min, and 3 min intervals and calculated their slopes. We statistically compared these slope values to values derived from the last 3-min of the 7-min trial, our 'gold' standard comparison. We found that a minimum of 2 min is required to achieve steady-rate metabolism and that data from 2-4 min yields a net COT curve that is not statistically different from the one derived from experimental protocols that are generally accepted in the field.

Utility of Inter-subject Transfer Learning for Wearable-Sensor-Based Joint Torque Prediction Models

Generalizability between individuals and groups is often a significant hurdle in model development for human subjects research. In the domain of wearable-sensor-controlled exoskeleton devices, the ability to generalize models across subjects or fine-tune more general models to individual subjects is key to enabling widespread adoption of these technologies. Transfer learning techniques applied to machine learning models afford the ability to apply and investigate the viability and utility such knowledge-transfer scenarios. This paper investigates the utility of single- and multi-subject based parameter transfer on LSTM models trained for "sensor-to-joint torque" prediction tasks, with regards to task performance and computational resources required for network training. We find that parameter transfer between both single- and multi-subject models provide useful knowledge transfer, with varying results across specific "source" and "target" subject pairings. This could be leveraged to lower model training time or computational cost in compute-constrained environments or, with further study to understand causal factors of the observed variance in performance across source and target pairings, to minimize data collection and model retraining requirements to select and personalize a generic model for personalized wearable-sensor-based joint torque prediction technologies.

A Neural Network Estimation of Ankle Torques From Electromyography and Accelerometry

Estimations of human joint torques can provide clinically valuable information to inform patient care, plan therapy, and assess the design of wearable robotic devices. Predicting joint torques into the future can also be useful for anticipatory robot control design. In this work, we present a method of mapping joint torque estimates and sequences of torque predictions from motion capture and ground reaction forces to wearable sensor data using several modern types of neural networks. We use dense feedforward, convolutional, neural ordinary differential equation, and long short-term memory neural networks to learn the mapping for ankle plantarflexion and dorsiflexion torque during standing, walking, running, and sprinting, and consider both single-point torque estimation, as well as the prediction of a sequence of future torques. Our results show that long short-term memory neural networks, which consider incoming data sequentially, outperform dense feedforward, neural ordinary differential equation networks, and convolutional neural networks. Predictions of future ankle torques up to 0.4 s ahead also showed strong positive correlations with the actual torques. The proposed method relies on learning from a motion capture dataset, but once the model is built, the method uses wearable sensors that enable torque estimation without the motion capture data.

Sensing leg movement enhances wearable monitoring of energy expenditure

Physical inactivity is the fourth leading cause of global mortality. Health organizations have requested a tool to objectively measure physical activity. Respirometry and doubly labeled water accurately estimate energy expenditure, but are infeasible for everyday use. Smartwatches are portable, but have significant errors. Existing wearable methods poorly estimate time-varying activity, which comprises 40% of daily steps. Here, we present a Wearable System that estimates metabolic energy expenditure in real-time during common steady-state and time-varying activities with substantially lower error than state-of-the-art methods. We perform experiments to select sensors, collect training data, and validate the Wearable System with new subjects and new conditions for walking, running, stair climbing, and biking. The Wearable System uses inertial measurement units worn on the shank and thigh as they distinguish lower-limb activity better than wrist or trunk kinematics and converge more quickly than physiological signals. When evaluated with a diverse group of new subjects, the Wearable System has a cumulative error of 13% across common activities, significantly less than 42% for a smartwatch and 44% for an activity-specific smartwatch. This approach enables accurate physical activity monitoring which could enable new energy balance systems for weight management or large-scale activity monitoring.

The exoskeleton expansion: improving walking and running economy

Since the early 2000s, researchers have been trying to develop lower-limb exoskeletons that augment human mobility by reducing the metabolic cost of walking and running versus without a device. In 2013, researchers finally broke this 'metabolic cost barrier'. We analyzed the literature through December 2019, and identified 23 studies that demonstrate exoskeleton designs that improved human walking and running economy beyond capable without a device. Here, we reviewed these studies and highlighted key innovations and techniques that enabled these devices to surpass the metabolic cost barrier and steadily improve user walking and running economy from 2013 to nearly 2020. These studies include, physiologically-informed targeting of lower-limb joints; use of off-board actuators to rapidly prototype exoskeleton controllers; mechatronic designs of both active and passive systems; and a renewed focus on human-exoskeleton interface design. Lastly, we highlight emerging trends that we anticipate will further augment wearable-device performance and pose the next grand challenges facing exoskeleton technology for augmenting human mobility.