Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Dec 1.
Published in final edited form as: Curr Opin Biomed Eng. 2023 Jul 24;28:100492. doi: 10.1016/j.cobme.2023.100492

Wearable sensing for understanding and influencing human movement in ecological contexts

Peter Gabriel Adamczyk a,*, Sara E Harper b, Alex J Reiter a, Rebecca A Roembke a, Yisen Wang a, Kieran M Nichols a, Darryl G Thelen a,b
PMCID: PMC10469849  NIHMSID: NIHMS1926467  PMID: 37663049

Abstract

Wearable sensors offer a unique opportunity to study movement in ecological contexts – that is, outside the laboratory where movement happens in ordinary life. This article discusses the purpose, means, and impact of using wearable sensors to assess movement context, kinematics, and kinetics during locomotion, and how this information can be used to better understand and influence movement. We outline the types of information wearable sensors can gather and highlight recent developments in sensor technology, data analysis, and applications. We close with a vision for important future research and key questions the field will need to address to bring the potential benefits of wearable sensing to fruition.

Keywords: wearable sensors, biomechanics, locomotion

1. Introduction

Wearables are transforming the field of biomechanics, with new technologies providing the capacity to measure the kinematics and kinetics of movement in an ever-widening array of out-of-lab contexts. These technologies range from now-common systems like wearable inertial sensors and pressure insoles to emerging ones like sonomyography and shear wave tensiometry. The tremendous quantity of wearable data, generated across a variety of conditions, requires new ways to process, analyze, and understand the information. This rapid development also provides an opportunity to reenvision what it means to study movement mechanics, especially how to investigate motor control, performance, and tissue loading in ecological contexts (Figure 1).

Figure 1:

Figure 1:

Open in a new tab

Uses of wearable sensors in ecological contexts. A: Location sensors, to determine context. B: Kinematic sensors to measure joint movements. C: Muscle and kinetic sensors to assess muscle activation patterns and loads on the body. D: Feedback and/or direct action to influence movement.

One key question in using biomechanical measurements to understand and influence movement in ecological contexts is “why?”. What drives the scientific community to do experiments that are less controlled and generate vast amounts of unstructured data, making it more challenging to interpret? What can be learned from implementing science in these contexts that cannot be learned in a lab? Collecting movement data outside the laboratory is motivated by the goal of investigating naturally emergent behavior, across a variety of conditions and behavioral states, and potentially with greater convenience. Wearable sensing can also be used to influence movement in real-world conditions, either indirectly through biofeedback, or directly as input to a real-time intervention. Wearable sensing enables a data-informed development cycle for these techniques and technologies, and real-world testing can ensure practical viability.

What is an “ecological context”, anyway? As in all science, it is axiomatic that any effort to study a system will disturb that system in some way. The term “ecological context” is about continual improvement in how well experimental conditions approximate the “real-life” conditions that motivate a particular study. Thus, there is a topic-dependent continuum; considering the example of a walking study, perhaps the least ecological context is a motor imagery study in fMRI (functional magnetic resonance imaging), and more ecological contexts may progress to treadmill walking, walking throughout a building, walking outdoors, and, ultimately, unsupervised walking in everyday life. These contexts vary in their level of similarity to everyday life, level of experimental control, sensors that can be applied, duration of recording, and even how well the context itself can be documented. Future directions for both wearable sensor technology and data analysis techniques should be formed by considering these aspects of ecological contexts in light of what we, as a field, want to learn and what we want to influence. This review will focus on these questions specifically in relation to lower-body kinematics and kinetics in locomotion-related tasks.

2. What do wearables help us understand?

2.1. How and where the body moves in space

One of the most important aspects of understanding and interpreting movement measurements is knowing the context in which a movement was performed. Context includes many factors that can affect movement, such as: surroundings, terrain, path constraints, weather, time of day, movement purpose, motivation level, physiological state, and more. Our own team has emphasized the importance of focusing on comparable contexts when aggregating data over long periods of time, especially specific paths that are traversed repeatedly [1] (Figure 2). The most common way of generating data about this context is through GPS location tracking. Because GPS does not work in buildings, the indoor part of the movement is usually lumped together. To improve on that, wearable inertial measurement units (IMUs) have been used with pedestrian dead-reckoning techniques to reconstruct location within buildings [1], and localization using other available signals is also under investigation [2]. Future progress could be made by combining sub-centimeter real-time kinematic (RTK) GPS with scanning and vision to achieve simultaneous localization and mapping (SLAM). If there is need for terrain information, the situation becomes more difficult, as it is rarely available in maps at the precision needed for investigating legged locomotion. As an alternative, efforts have mounted scanners or cameras to the body to identify features like stairs [3,4] or even the relationship between viable footholds and gaze [5–7]. Other aspects of context, such as purpose or motivation, can be even more difficult. It may be necessary to glean information from test participants themselves, such as through voice recordings that report the salient context of interesting movements or events; for example, losses of balance [8–10].

Figure 2:

Figure 2:

Figure 2:

Open in a new tab

Reducing gait data from ecological contexts to focus on frequently-repeated paths. Left: Conceptual diagram showing the reduction of variability by focusing analysis only on locations and paths that are used most frequently. Right: Example result (modified from [1] with permission) showing that eliminating data from other paths reduced variability in comparison to the set of all straight-line walking.

2.2. How the joints move the body

In addition to the movement context discussed above, biomechanics focuses on analysis of joint-level movements. The critical aspect of the ecological context is that it is uncontrolled – for example, it may include a range of conditions that are irreproducible or unknown. Critical issues in understanding joint movement in ecological contexts include how to measure the movement, how to process and analyze the data, and how to combine it with contextual information and interpret it to generate meaningful knowledge.

The motion of the body in ecological contexts is now commonly measured using inertial measurement units (IMUs). Raw IMU measurements can be processed through sensor-fusion algorithms and kinematic models of the body to estimate a wide variety of movement parameters in a variety of ways. Advancements continue to be made in estimating and using joint kinematics [11–16], foot movement [17–19], center of mass mechanics [20,21] and other aspects of movement. Another novel way to characterize segment and whole body movements is using tactile textiles with piezoresistive fibers, interpreted through machine learning techniques [22]. Beyond the technical reconstruction of movement, an outstanding challenge of understanding wearable studies in ecological contexts lies in determining which data are meaningful for answering a given question. Here it is critical for researchers to choose instrumentation and data processing that can feasibly answer the questions under study. This means ensuring sufficient accuracy and precision in the measurements, including awareness of the strengths and limitations of each sensor type and algorithm.

2.3. How the muscles move the joints

Kinematic characterization of movement is valuable, but deeper understanding is attained by knowing the control and mechanics of the muscles that generate and sustain the movement. As with kinematic analysis, the ability to measure musculoskeletal kinetics in ecological contexts is important because these contexts generate unusual scenarios that are uniquely relevant to injury, extremes of performance, and daily life. The key issue in kinetic analysis with wearables is how to assess load and power information without stationary force sensors such as in-ground force plates.

One approach to force sensing has been to incorporate miniature force plates in shoes; this can work, but it creates a bulky and stiff sole, which changes ground contact mechanics [23]. A common alternative is pressure insoles, which can be made thin and flexible [24,25]; however, current solutions provide only a subset of foot-floor reactions. A full solution must also measure shear stress, and this remains an unresolved area of research (e.g. [26]). To circumvent the limitations of these systems, modeling approaches have been proposed to estimate joint kinetics directly from movement kinematics [16,27,28] or from kinematics and sparse ground reactions [25]. In general, this still requires great expertise in the design, application, and interpretation of such inverse models. Further, tremendous amounts of data will be needed to make these approaches generalizable for diverse populations and ecological conditions.

A different approach to understanding movement kinetics is to measure, or estimate, the state of the body’s tissues more directly. The most straightforward technology is electromyography (EMG). Muscle modeling approaches can attempt to use EMG data to estimate muscle forces [29,30]; however, the relationship between EMG signals and muscle forces can change over time, due to both intrinsic muscle state (e.g. length, velocity, fatigue) and extrinsic factors such as changes in skin impedance or sensor placement. Approaches using dense multi-channel EMG arrays [31,32] can reduce some of these limitations to better study muscle function with robustness to sensor placement error and the ability to decompose signals into fine motor units. Recently, shear wave tensiometry (SWT) has emerged to measure axial muscle-tendon loading via the speed of a wave propagating along the taut tendon [33]; SWT has been used to detect changes in tissue load in response to variable external conditions such as terrain slope or exosuit assistance [34,35]. Other recent approaches aim to use advances in portable ultrasound to understand both kinematic and kinetic muscle function [36]. Sonomyography uses ultrasound images of either a single depth line or a planar field, tracking features such as muscle thickness using computer vision techniques, to estimate kinematics and/or force [37,38] or classify movement [39,40].

Some emerging alternatives for assessing muscle state are worth considering as well. One is mechanomyography (MMG), a technique of measuring the low-frequency vibrations of contracting muscles, allowing for movement pattern recognition and subsequent biofeedback [41]. Another, different concept, coincidentally with a similar name, is surface mechanomyography, which uses stretch sensors wrapped over the skin to measure the bulging of muscles during movement [42]. While little-evaluated to date, these sensors are low-profile enough to potentially be mobile for ecological contexts. Finally, a recent novel technique focuses on joint contact forces rather than muscle loads, analyzing spontaneous acoustic emissions from joint articular surfaces (e.g. “knee sounds”) together with machine learning to estimate joint loading during movement [43]. If proven reliable and repeatable, such an approach could address important questions about internal joint loads that are relevant to assess injury risk, while also providing valuable data to tune and validate musculoskeletal models.

One of the major challenges in understanding muscle kinetics in ecological contexts is to get sensors that are wearable enough to be part of an experiment beyond the lab. Most of the sensors described above are themselves small enough to be wearable, but few are readily packaged for ecological use. Making more of these systems small, lightweight, battery-powered, and with adequate portable data logging and synchronization capabilities would enable richer studies of how the body’s tissues are loaded in the variable contexts of real-world activities. Another significant opportunity is to combine these systems together to answer questions about muscle-tendon energetics and mechanics during ecological movement (e.g. tensiometry and IMUs [44]; EMG and IMUs [30]; EMG and sonomyography [38,45,46]; MMG and IMUs [41]).

3. How can wearable sensing influence movement?

The measurements above of location, motion, and kinetics are keys to understanding ecological movement. The next step is to apply these technologies to influence movement, not just observe it. Three basic methods are available: using summary information to drive high-level behavioral change; using instantaneous measurements to change low-level movement patterns through biofeedback; and incorporating measurement technology into wearable devices that directly alter movement.

Driving behavioral change is the mechanism used by consumer wearables such as smart watches, rings or shoe pods, and some sports-related devices. These systems monitor movement quantity and sometimes detailed characteristics (e.g. cadence, impact acceleration), and provide summary data to help the user make decisions, such as whether to get more or less exercise, or to show whether performance is improving or degrading over time. Most of the wearable measurement systems can generate behavior-modifying information, so the question for each of them is how best to relate the specific signals to behavioral recommendations.

Biofeedback is a very intensive use of wearable sensor information, as it requires real-time processing and display to the user in a format that can be immediately acted on to modify a movement while it is happening. Because of the demands of real-time display, biofeedback applications are most suitable for raw or near-raw signals (to minimize processing latency) that are easily understood by a user. Examples include filtered EMG or sonomyography, single-joint kinematics from IMUs, or foot contact pressure, on one limb (for targeted improvements) or bilaterally (for symmetry). As embedded processing power increases, more sensors may become real-time capable, and ways of displaying information may advance from current screen-based, vibrotactile, or audible signals to advanced displays like augmented reality (see Figure 1D).

Finally, the incorporation of real-time measurements in devices that directly influence movement is an exciting present and near-future domain. In this case, the bandwidth problem of conveying information to a human user is circumvented; but it is replaced with a problem of intelligently using the measured information to control a device. Examples include exoskeletons and prostheses that respond to movement, EMG, or embedded force measurement signals [47–50], or electrical stimulation devices that modulate timing and/or amplitude in synchrony with measured movement [51–54]. A current frontier in this work is in truly bionic devices that include bidirectional information transfer as well as direct control or influence over movement in parallel with a human user [55].

4. Vision: Future Priorities and Philosophical Questions

The many current and potential technologies for field-based wearable instrumentation lead to the question:What should be the priorities for near-term research? Several areas of critical need are outlined here:

  • Deep tissue measurements: Most current wearable muscle or tendon measurements are most sensitive to and representative of the superficial tissues they are near. For fuller understanding, it will remain valuable to continue developing technologies to look deeper into the body during movement.

  • Novel phenomena: The range of sensors that can potentially be made wearable opens a wealth of opportunities to measure phenomena that could not previously be observed. By addressing questions of movement and loading from multiple different angles, new richness could be added to mechanistic studies and to advanced modeling techniques.

  • Convenience: Battery life, automatic record/sleep modes, cleanability, error recovery, and other issues that burden the user need to be improved if wearable sensing is to achieve continual use.

  • Data quality and robustness: Using wearable systems in ecological contexts challenges the technology substantially, in areas from limited computing power and communication, to severe movement and environmental conditions, and even to use with varied and conflicting equipment (e.g. boots, sports equipment, outerwear, tools). It is critical to make sensing systems easily deployable and robust, and also to understand their limitations when choosing what to use in a given study.

  • Data standardization and reduction: Modern science and industry draw great value from data-driven applications, but variable, poorly curated, or uninteresting data are a major challenge to these endeavors. Researchers and practitioners should take great care to define standard procedures for wearable technologies, so that data representing different ecological contexts is sure to represent the context, and not just variable test protocols.

  • Statistical and interpretive approaches: Especially in long-term studies that involve multiple types of movements, simple statistics cannot justly describe the recorded data. Continued development or adaptation of richer statistical interpretive tools, such as whole-sample histogram analysis, or classification-based or cluster-based subset analysis and other machine learning approaches, will be very important for generating meaning from the rich data sets collected.

In addition to these practical concerns, we see several overarching philosophical questions the field should address:

  • Richness vs. simplicity: The many potential options available make it tempting to heavily instrument test participants, but even wearable instrumentation can become cumbersome in quantity. Researchers should be cognizant of the trade-off between data richness (from many sensors) and true ecological validity (few enough sensors to forget about them).

  • Typical signals vs. outlier events: Wearable data from ecological contexts is likely to capture a great deal of common movement, plus a few strange and unique events. Both categories are interesting: the first, to understand typical movement patterns and coordination; and the second to study real-life occurrences that are extremely rare in the lab, such as falls or injurious movements. For greatest insight, researchers should be attentive to finding and interpreting both (see Figure 3).

  • Understanding variability vs. rejecting variability: Even within typical movement, there is variation, and some aspects of the variation are important while others may not be. Researchers should attend to methods of understanding why different types of variability arise, but should also attend to improving methods for obtaining the most accurate measures of “typical movement” by filtering irrelevant variability.

  • Usability vs. insight: Many of the most successful recent advancements include machine learning in at least some component of data analysis or device control. While these techniques are powerful, the common complaint is that the underlying models are usually uninterpretable, severely limiting the insight that can be gleaned. As in other fields, researchers should advance on both fronts – enabling important applications while still creating generalizable knowledge.

  • Privacy vs. data sharing: With better resolution, improving data richness, and longer-term real-world collection come privacy risks that are not prevalent in current wearable data sets. In particular, context information (e.g. location, time of day) is enough to disclose identity and lifestyle choices. The legal, ethical, and practical management of privacy while making the most of each data set is an area worthy of dedicated attention and standardization in the field.

Figure 3:

Figure 3:

Open in a new tab

Information flow to generate insight about common or unique movements from wearable sensor data.

5. Conclusion

A variety of traditional and emerging technologies are enabling new ways of studying movement in ecological contexts, outside the laboratory. With these capabilities come challenges with respect to the sensors themselves, including interfacing, processing, statistical and interpretation approaches; and to the scientific and philosophical underpinnings of these approaches. A rich variety of research endeavors are motivated, to enable a future where biomechanical data have maximal meaning and relevance to solving problems and improving lives.

Acknowledgments

Funding:

The authors’ work related to this article was supported by the U.S. National Institutes of Health [grant AR074897 and TL1TR002375], the U.S. Department of Defense [grants W81XWH-19-2-0024, W81XWH-20-1-0884, and W81XWJ-21-2-0006], and the National Science Foundation [grant CBET-2019621 and DGE-1747503] for supporting relevant portions of our work. Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessary reflect the views of any funding agency. The authors thank Dylan Schmitz for helpful comments and Adrienne Nienow for preparing the graphics in Figure 2.

Peter G. Adamczyk reports financial support was provided by National Institutes of Health. Darryl G. Thelen reports financial support was provided by National Institutes of Health. Darryl G. Thelen reports financial support was provided by US Department of Defense. Peter G. Adamczyk reports financial support was provided by US Department of Defense. Darryl G. Thelen has patent issued to Wisconsin Alumni Research Foundation. Peter G. Adamczyk has patent pending to Wisconsin Alumni Research Foundation. Alex J. Reiter has patent pending to Wisconsin Alumni Research Foundation. Sara E Harper has patent pending to Wisconsin Alumni Research Foundation.

Footnotes

Wearable sensors enable study of human movement in real-world “ecological” settings.

Sensed signals include location, body kinematics, muscle activity and load, and more.

Advancements include novel sensors, data analysis, interpretation, and combinations.

Sensors can influence movement through behavioral changes, biofeedback, and devices.

A vision of future priorities and philosophical questions is introduced.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

Papers of particular interest, published within the period of review, have been highlighted as: * of special interest

  • 1. Wang W, Adamczyk PG: Analyzing Gait in the Real World Using Wearable Movement Sensors and Frequently Repeated Movement Paths. Sensors 2019, 19:1925. * Work by author(s) of the current article. An improved method to reconstruct detailed path and foot movement from an inertial sensor, a GPS receiver and a barometric altitude sensor using strapdown navigation and sensor fusion algorithms during days-long measurements. Also introduces an analysis algorithm focused on detecting frequently-repeated straight walking paths that reduce variability to allow for proper statistical comparisons of gait for everyday, natural movement behavior.
  • 2.Obeidat H, Shuaieb W, Obeidat O, Abd-Alhameed R: A Review of Indoor Localization Techniques and Wireless Technologies. Wireless Pers Commun 2021, 119:289–327. [Google Scholar]
  • 3.Diaz JP, Silva RL da, Zhong B, Huang HH, Lobaton E: Visual Terrain Identification and Surface Inclination Estimation for Improving Human Locomotion with a Lower-Limb Prosthetic. In 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). . 2018:1817–1820. [DOI] [PubMed] [Google Scholar]
  • 4.Li M, Zhong B, Lobaton E, Huang H: Fusion of Human Gaze and Machine Vision for Predicting Intended Locomotion Mode. IEEE Transactions on Neural Systems and Rehabilitation Engineering 2022, 30:1103–1112. [DOI] [PubMed] [Google Scholar]
  • 5.Ghiani A, Van Hout LR, Driessen JG, Brenner E: Where do people look when walking up and down familiar staircases? Journal of Vision 2023, 23:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Wirth T, Matthis J: The relationship between gaze and foot placement is shaped by the visual discriminability and availability of footholds in an overground Augmented Reality stepping stone task. Journal of Vision 2022, 22:4033. [Google Scholar]
  • 7.Matthis JS, Yates JL, Hayhoe MM: Gaze and the Control of Foot Placement When Walking in Natural Terrain. Current Biology 2018, 28:1224–1233.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Ojeda LV, Adamczyk PG, Rebula JR, Nyquist LV, Strasburg DM, Alexander NB: Reconstruction of body motion during self-reported losses of balance in community-dwelling older adults. Medical Engineering & Physics 2019, 64:86–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Hauth J, Jabri S, Kamran F, Feleke EW, Nigusie K, Ojeda LV, Handelzalts S, Nyquist L, Alexander NB, Huan X, et al. : Automated Loss-of-Balance Event Identification in Older Adults at Risk of Falls during Real-World Walking Using Wearable Inertial Measurement Units. Sensors 2021, 21:4661. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Nouredanesh M, Ojeda L, Alexander NB, Godfrey A, Schwenk M, Melek W, Tung J: Automated Detection of Older Adults’ Naturally-Occurring Compensatory Balance Reactions: Translation From Laboratory to Free-Living Conditions. IEEE Journal of Translational Engineering in Health and Medicine 2022, 10:1–13. [Google Scholar]
  • 11.Potter MV, Cain SM, Ojeda LV, Gurchiek RD, McGinnis RS, Perkins NC: Evaluation of Error-State Kalman Filter Method for Estimating Human Lower-Limb Kinematics during Various Walking Gaits. Sensors 2022, 22:8398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Weygers I, Kok M, De Vroey H, Verbeerst T, Versteyhe M, Hallez H, Claeys K: Drift-Free Inertial Sensor-Based Joint Kinematics for Long-Term Arbitrary Movements. IEEE Sensors Journal 2020, 20:7969–7979. [Google Scholar]
  • 13.Weygers I, Kok M, Konings M, Hallez H, De Vroey H, Claeys K: Inertial Sensor-Based Lower Limb Joint Kinematics: A Methodological Systematic Review. Sensors 2020, 20:673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.McGrath T, Stirling L: Body-Worn IMU-Based Human Hip and Knee Kinematics Estimation during Treadmill Walking. Sensors 2022, 22:2544. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Chen Y, Fu C, Leung WSW, Shi L: Drift-Free and Self-Aligned IMU-Based Human Gait Tracking System With Augmented Precision and Robustness. IEEE Robotics and Automation Letters 2020, 5:4671–4678. [Google Scholar]
  • 16.Mundt M, Johnson WR, Potthast W, Markert B, Mian A, Alderson J: A Comparison of Three Neural Network Approaches for Estimating Joint Angles and Moments from Inertial Measurement Units. Sensors 2021, 21:4535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Chen X, Zhang K, Liu H, Leng Y, Fu C: A Probability Distribution Model-Based Approach for Foot Placement Prediction in the Early Swing Phase With a Wearable IMU Sensor. IEEE Transactions on Neural Systems and Rehabilitation Engineering 2021, 29:2595–2604. [DOI] [PubMed] [Google Scholar]
  • 18.Baroudi L, Newman MW, Jackson EA, Barton K, Shorter KA, Cain SM: Estimating Walking Speed in the Wild. Frontiers in Sports and Active Living 2020, 2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Baroudi L, Yan X, Newman MW, Barton K, Cain SM, Shorter KA: Investigating walking speed variability of young adults in the real world. Gait & Posture 2022, 98:69–77. * Two weeks of continuous data from a thigh-worn accelerometer and a heart rate monitor identified that the continuity and duration of a walk can be used to explain some of the variability in real-world walking speed. Speeds estimated from long continuous walks had the lowest standard deviation; these portions should be used to confidently isolate the preferred walking behavior of an individual. Also presents valuable ways of showing and comparing statistical distributions of large data sets.
  • 20.Slaughter PR, Adamczyk PG: Tracking Quantitative Characteristics of Cutting Maneuvers with Wearable Movement Sensors during Competitive Women’s Ultimate Frisbee Games. Sensors 2020, 20:6508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lee M, Park S: Estimation of Three-Dimensional Lower Limb Kinetics Data during Walking Using Machine Learning from a Single IMU Attached to the Sacrum. Sensors 2020, 20:6277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Luo Y, Li Y, Sharma P, Shou W, Wu K, Foshey M, Li B, Palacios T, Torralba A, Matusik W: Learning human–environment interactions using conformal tactile textiles. Nat Electron 2021, 4:193–201. [Google Scholar]
  • 23.Santos JP, Ferreira JP, Crisóstomo M, Paulo Coimbra A: Instrumented Shoes for 3D GRF Analysis and Characterization of Human Gait. In Bioinformatics and Biomedical Engineering. Edited by Rojas I, Valenzuela O, Rojas F, Ortuño F. Springer International Publishing; 2019:51–62. [Google Scholar]
  • 24.Cramer LA, Wimmer MA, Malloy P, O’Keefe JA, Knowlton CB, Ferrigno C: Validity and Reliability of the Insole3 Instrumented Shoe Insole for Ground Reaction Force Measurement during Walking and Running. Sensors 2022, 22:2203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Hullfish TJ, Baxter JR: A simple instrumented insole algorithm to estimate plantar flexion moments. Gait & Posture 2020, 79:92–95. [DOI] [PubMed] [Google Scholar]
  • 26.Fernandes J, Chen J, Jiang H: Three-Axis Capacitive Sensor Arrays for Local and Global Shear Force Detection. Journal of Microelectromechanical Systems 2021, 30:799–813. [Google Scholar]
  • 27. Al Borno M, O’Day J, Ibarra V, Dunne J, Seth A, Habib A, Ong C, Hicks J, Uhlrich S, Delp S: OpenSense: An open-source toolbox for inertial-measurement-unit-based measurement of lower extremity kinematics over long durations. Journal of NeuroEngineering and Rehabilitation 2022, 19:22. * An open-source tool and workflow using sensor fusion and an inverse kinematics approach with a constrained biomechanical model, to produce joint kinematics from IMUs consistent with those estimated by optical motion capture. Importantly, the workflow is capable of assessing and mitigating IMU kinematic drift, potentially enabling long-duration measurements.
  • 28.Karatsidis A, Jung M, Schepers HM, Bellusci G, de Zee M, Veltink PH, Andersen MS: Musculoskeletal model-based inverse dynamic analysis under ambulatory conditions using inertial motion capture. Medical Engineering & Physics 2019, 65:68–77. [DOI] [PubMed] [Google Scholar]
  • 29.Siu HC, Sloboda J, McKindles RJ, Stirling LA: A Neural Network Estimation of Ankle Torques From Electromyography and Accelerometry. IEEE Transactions on Neural Systems and Rehabilitation Engineering 2021, 29:1624–1633. [DOI] [PubMed] [Google Scholar]
  • 30.Gurchiek RD, Donahue N, Fiorentino NM, McGinnis RS: Wearables-Only Analysis of Muscle and Joint Mechanics: An EMG-Driven Approach. IEEE Transactions on Biomedical Engineering 2022, 69:580–589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Gallina A, Disselhorst-Klug C, Farina D, Merletti R, Besomi M, Holobar A, Enoka RM, Hug F, Falla D, Søgaard K, et al. : Consensus for experimental design in electromyography (CEDE) project: High-density surface electromyography matrix. Journal of Electromyography and Kinesiology 2022, 64:102656. * Recommendations on the use of high-density surface electromyography (HDsEMG) in experimental studies developed by the Consensus for Experimental Design in Electromyography (CEDE) project. The provided matrix is intended to help researchers when collecting, reporting, and interpreting HDsEMG data.
  • 32.Caillet AH, Avrillon S, Kundu A, Yu T, Phillips ATM, Modenese L, Farina D: Larger and denser: an optimal design for surface grids of EMG electrodes to identify greater and more representative samples of motor units. 2023, doi: 10.1101/2023.02.18.529050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Martin JA, Brandon SCE, Keuler EM, Hermus JR, Ehlers AC, Segalman DJ, Allen MS, Thelen DG: Gauging force by tapping tendons. Nature Communications 2018, 9:1592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Harper SE, Roembke RA, Zunker JD, Thelen DG, Adamczyk PG: Wearable Tendon Kinetics. Sensors 2020, 20:4805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Schmitz DG, Thelen DG, Cone SG: A Kalman Filter Approach for Estimating Tendon Wave Speed from Skin-Mounted Accelerometers. Sensors 2022, 22:2283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.de Oliveira J, de Souza MA, Assef AA, Maia JM: Multi-Sensing Techniques with Ultrasound for Musculoskeletal Assessment: A Review. Sensors 2022, 22:9232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Hallock LA, Sud B, Mitchell C, Hu E, Ahamed F, Velu A, Schwartz A, Bajcsy R: Toward Real-Time Muscle Force Inference and Device Control via Optical-Flow-Tracked Muscle Deformation. IEEE Transactions on Neural Systems and Rehabilitation Engineering 2021, 29:2625–2634. * Demonstrates the potential utility of muscle deformation for both biomechanical study of individual muscle dynamics and device control. The technique is readily extensible to multiple muscles and device degrees of freedom.
  • 38.Jahanandish MH, Fey NP, Hoyt K: Lower Limb Motion Estimation Using Ultrasound Imaging: A Framework for Assistive Device Control. IEEE Journal of Biomedical and Health Informatics 2019, 23:2505–2514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Engdahl SM, Acuña SA, King EL, Bashatah A, Sikdar S: First Demonstration of Functional Task Performance Using a Sonomyographic Prosthesis: A Case Study. Front Bioeng Biotechnol 2022, 10:876836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Engdahl S, Dhawan A, Bashatah A, Diao G, Mukherjee B, Monroe B, Holley R, Sikdar S: Classification Performance and Feature Space Characteristics in Individuals With Upper Limb Loss Using Sonomyography. IEEE Journal of Translational Engineering in Health and Medicine 2022, 10:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Paszkiewicz FP, Wilson S, Oddsson M, McGregor AH, Alexandersson Á, Huo W, Vaidyanathan R: Microphone Mechanomyography Sensors for Movement Analysis and Identification. In 2022. International Conference on Advanced Robotics and Mechatronics (ICARM). . 2022:118–125. [Google Scholar]
  • 42.Surface Mechanomyography (sMMG) Sensor Technology Explained - Figur8. 2021, [Google Scholar]
  • 43. Scherpereel KL, Bolus NB, Jeong HK, Inan OT, Young AJ: Estimating Knee Joint Load Using Acoustic Emissions During Ambulation. Ann Biomed Eng 2021, 49:1000–1011. * Introduces a novel technique demonstrating the sufficiency of using joint acoustic emissions for estimating internal joint load in wearable applications. In estimating joint contact force under varying conditions, subject-specific models using features of the joint acoustics achieved lower errors than models based on traditional gait measures of EMG, GRF, and motion capture.
  • 44. Harper SE, Schmitz DG, Adamczyk PG, Thelen DG: Fusion of Wearable Kinetic and Kinematic Sensors to Estimate Triceps Surae Work during Outdoor Locomotion on Slopes. Sensors 2022, 22:1589. * Work by author(s) of the current article. The introduction of a wearable system combining shear-wave tensiometry (SWT) and inertial measurement units (IMUs) for the analysis of muscle-tendon power output. Paired estimates of muscle force and length changes yield step-by-step calculations of work output which vary with continuous variables - here, ground incline. This portable system enables monitoring of kinetic, kinematic, and combined metrics of muscle performance in the field.
  • 45.Rabe KG, Fey NP: Evaluating Electromyography and Sonomyography Sensor Fusion to Estimate Lower-Limb Kinematics Using Gaussian Process Regression. Front Robot AI 2022, 9:716545. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Zhang Q, Fragnito N, Franz JR, Sharma N: Fused ultrasound and electromyography-driven neuromuscular model to improve plantarflexion moment prediction across walking speeds. J NeuroEngineering Rehabil 2022, 19:86. * The development of a predictive model of plantarflexion moment, pairing surface EMG (sEMG) with ultrasound (US) to improve prediction accuracy. The incorporation of US accounts for dynamic muscle movement in addition to the muscle activation captured by sEMG. Combining these modalities to drive the model and calibrating it across task conditions significantly reduces prediction error, showing promise for use in bio-inspired control strategies for rehabilitative devices.
  • 47.Slade P, Kochenderfer MJ, Delp SL, Collins SH: Personalizing exoskeleton assistance while walking in the real world. Nature 2022, 610:277–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Conner BC, Lerner ZF: Improving Ankle Muscle Recruitment via Plantar Pressure Biofeedback during Robot Resisted Gait Training in Cerebral Palsy. In 2022. International Conference on Rehabilitation Robotics (ICORR). . 2022:1–6. * Validation of wearable exoskeleton device for gait training in children with cerebral palsy, providing plantar pressure biofeedback to facilitate user engagement with plantar flexor robotic resistance. The relatively simple addition of plantar pressure feedback prompts increased soleus activity and demonstrates viability for clinical implementation.
  • 49.Tran M, Gabert L, Hood S, Lenzi T: A lightweight robotic leg prosthesis replicating the biomechanics of the knee, ankle, and toe joint. Sci Robot 2022, 7:eabo3996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Farina D, Burdet E, Mehring C, Ibáñez J: Roboticists Want to Give You a Third Arm: Unused Bandwidth in Neurons Can be Tapped to Control Extra Limbs. IEEE Spectrum 2023, 60:22–46. * Explores the use of EMG arrays to harness unused neural signals to control robotic limbs. While completing a motor task of their tibialis anterior muscle, participants channeled extraneous neural frequencies to control cursor movement about a monitor. Further studies will investigate how high frequency components of neural signals, unused in muscle control, can be used to control additional robotic activities.
  • 51.CIONIC: superpowering the human body. [date unknown],
  • 52.Robison J, Gibbons R, Achelis D, Bent B, Wajda D, Webster R: Augmenting gait in a population exhibiting foot drop with adaptive functional electrical stimulation. 2022, doi: 10.1101/2022.04.27.22273623. [DOI] [Google Scholar]
  • 53.Kahya M, Hackman D, Jacobs L, Nilsson D, Rumsey Y, Oddsson LIE: Wearable Technologies Using Peripheral Neuromodulation to Enhance Mobility and Gait Function in Older Adults―A Narrative Review. The Journals of Gerontology: Series A 2022, doi: 10.1093/gerona/glac045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Thorp JE, Adamczyk PG: Mechanisms of gait phase entrainment in healthy subjects during rhythmic electrical stimulation of the medial gastrocnemius. PLOS ONE 2020, 15:e0241339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55. Pasluosta C, Kiele P, Čvančara P, Micera S, Aszmann OC, Stieglitz T: Bidirectional bionic limbs: a perspective bridging technology and physiology. J Neural Eng 2022, 19:013001. * A perspective article proposing collaborative efforts to combine technological advancements in bidirectionally communicative bionic limbs with current knowledge of amputation pathophysiology and motor control. Uniting these lines of research and capitalizing on the resulting insights may help to realize the next generation of bionic limbs.

RESOURCES