Simplified Markerless Stride Detection Pipeline (sMaSDP) for Surface EMG Segmentation
<p>Simplified stages of a typical gait cycle for a right leg dominant subject. A new cycle begins with the right heel strike (HS) and both feet in contact with the floor. After a loading response, the left foot is lifted off the floor following the left Toe-off (TO), and weight starts shifting on the right foot from heel to mid-foot as the body moves forward. Once the body reaches mid-stance, the right heel loses contact with the floor during the right Heel-off (HO) stage, and weight shifts towards the right forefoot. To prepare for the right limb swing phase, the left foot contacts the floor again with a left heel strike (HS), and both feet are in contact with the floor, which is called double support. The right foot then enters the swing phase following the right Toe-off (TO), and the same stance process happens on the left foot, with weight shifting from heel to toe, and the respective left Heel-off (HO). The current gait cycle ends with the right HS, and a new cycle starts.</p> "> Figure 2
<p>Schematic of the sEMG and IMU sensor placement locations. sEMG recordings were taken from Tibialis Anterior (TA), Medial Gastrocnemius (mGAST), Vastus Lateralis (VL), Rectus Femoris (RF), Semitendinosus (SEM) and Biceps Femoris Longus (BFL). The IMU sensor was attached on top of the subject’s shoe, at the centre of the right foot.</p> "> Figure 3
<p>Summary of method architecture: the diverse multi-step process, from the recorded IMU signals, filtering of movement periods, segmentation into different walking modalities and identification of HS moments. Steps labelled <b>A</b> to <b>E</b> establish a logical link between this architecture and the results.</p> "> Figure 4
<p>Summary diagram illustrating the different stages of the method, presented and linked in <a href="#sensors-23-04340-f003" class="html-fig">Figure 3</a>, for the duration of an HT (RD, LGW, SA): in Section (<b>A</b>), the <span class="html-italic">x</span>-axis acceleration of the right-foot IMU is used to identify the initial HT start and end timepoints, marked with the green circles; within the time range defined by the HT green circles, Section (<b>B</b>) now takes the <span class="html-italic">y</span>-axis position data from the IMU to identify the sharp direction turns related to switches in walking modality, represented here by the yellow triangles; lastly, within the isolated LGW modality between the yellow triangles, Section (<b>C</b>) uses the <span class="html-italic">z</span>-axis acceleration from the IMU, and through our filtering and peak identification algorithm, identifies the HS events, marked by the red triangles.</p> "> Figure 5
<p>Segmented sEMG from a single extracted gait cycle: top section diagram represents a full gait cycle, as initially described (<a href="#sensors-23-04340-f001" class="html-fig">Figure 1</a>), with the respective locations of the sEMG sensors; the bottom traces are the segmented sEMG recordings for the respective gait cycle. Linked, as labelled, to step <b>D</b> of the method architecture (<a href="#sensors-23-04340-f003" class="html-fig">Figure 3</a>).</p> "> Figure 6
<p>Observations with mean and standard deviation of muscle activity extracted from 12 LGW gait cycles from a randomly selected trial and subject. Analysing this activity in different walking modalities allows for quantification of the variability in each muscle and the impact on the person’s mobility. Linked, as labelled, to step <b>E</b> of the method architecture (<a href="#sensors-23-04340-f003" class="html-fig">Figure 3</a>).</p> "> Figure 7
<p>HS detection on unconstrained Parkinson gait, computed from IMU data from the publicly available dataset of [<a href="#B33-sensors-23-04340" class="html-bibr">33</a>,<a href="#B39-sensors-23-04340" class="html-bibr">39</a>]. The presented data were recorded from a PD patient, supervised by a therapist. Within this recording the PD patient walks freely in three separate modalities: straight walking, “random walking” and ADL walking. HS events are here detected for all modalities.</p> ">
Abstract
:1. Introduction
2. Methodology
- RA: ramp ascent.
- RD: ramp descent.
- SA: staircase ascent.
- SD: staircase descent.
- LGW: level ground walking.
2.1. Interpolation and Kinematic Activity Segmentation
2.2. Filtering Small Movement Artefacts and Separating Trials from a Single Recording
2.3. Initial sEMG Segmentation and Direction, Using the Timepoints Defined by the Four Moments of Activity per Complete Trial
2.4. Walking Modality Identification within Each Direction of the Course and Trials
2.5. Using Acceleration Data to Identify the Moment of Heel Strike
- the high-pass filter is applied to the raw KIN data, removing frequencies below the 20 Hz band, and potential movement artefacts.
- a half-wave rectifier is applied to the resulting filtered data, removing unnecessary negative electrical oscillations imposed by the sensor readings.
- a low-pass filter is applied to the rectified data, serving as an anti-aliasing filter at 5 Hz, to remove oscillations from the signal.
3. Results
4. Discussion and Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Abbreviations
ACC | Acceleration |
ADLs | Activities of Daily Living |
DOAJ | Directory of open access journals |
EMG | Electromyography |
EC | End-of-Contact |
FSR | Force Sensitive Resistors |
HS | Heel-strike |
HO | Heel-off |
HT | Half Trial |
IC | Initial Contact |
IPS | Insole Pressure Sensors |
IMU | Inertial Measurement Unit |
KIN | Kinematics |
LGW | Level Ground Walking |
MDPI | Multidisciplinary Digital Publishing Institute |
PD | Parkinson’s Disease |
POS | Position |
RA | Ramp Ascent |
RD | Ramp Descent |
SA | Stairs Ascent |
SD | Stairs Descent |
T | Trial |
TC | Terminal Contact |
TO | Toe-off |
TS | Time Stamps |
VEL | Velocity |
WM | Walking Modality |
References
- Whittle, M. Gait Analysis: An Introduction; Butterworth-Heinemann: Oxford, UK, 2007. [Google Scholar]
- Tao, W.; Liu, T.; Zheng, R.; Feng, H. Gait Analysis Using Wearable Sensors. Sensors 2012, 12, 2255–2283. [Google Scholar] [CrossRef] [PubMed]
- Gujarathi, T.; Bhole, K. Gait Analysis using IMU sensor. In Proceedings of the 2019 10th International Conference on Computing, Communication and Networking Technologies (ICCCNT), Kanpur, India, 6–8 July 2019; pp. 1–5. [Google Scholar] [CrossRef]
- Winter, D.A.; Greenlaw, R.K.; Hobson, D.A. A microswitch shoe for use in locomotion studies. J. Biomech. 1972, 5, 553–554. [Google Scholar] [CrossRef]
- Hanlon, M.; Anderson, R. Real-time gait event detection using wearable sensors. Gait Posture 2009, 30, 523–527. [Google Scholar] [CrossRef] [PubMed]
- Han, Y.C.; Wong, K.I.; Murray, I. Gait Phase Detection for Normal and Abnormal Gaits Using IMU. IEEE Sens. J. 2019, 19, 3439–3448. [Google Scholar] [CrossRef]
- Nazmi, N.; Abdul Rahman, M.A.; Yamamoto, S.I.; Ahmad, S.A. Walking gait event detection based on electromyography signals using artificial neural network. Biomed. Signal Process. Control 2019, 47, 334–343. [Google Scholar] [CrossRef]
- Prasanth, H.; Caban, M.; Keller, U.; Courtine, G.; Ijspeert, A.; Vallery, H.; von Zitzewitz, J. Wearable Sensor-Based Real-Time Gait Detection: A Systematic Review. Sensors 2021, 21, 2727. [Google Scholar] [CrossRef]
- Morris, J. Accelerometry—A technique for the measurement of human body movements. J. Biomech. 1973, 6, 729–736. [Google Scholar] [CrossRef] [PubMed]
- Jasiewicz, J.M.; Allum, J.H.; Middleton, J.W.; Barriskill, A.; Condie, P.; Purcell, B.; Li, R.C.T. Gait event detection using linear accelerometers or angular velocity transducers in able-bodied and spinal-cord injured individuals. Gait Posture 2006, 24, 502–509. [Google Scholar] [CrossRef]
- Tadano, S.; Takeda, R.; Miyagawa, H. Three Dimensional Gait Analysis Using Wearable Acceleration and Gyro Sensors Based on Quaternion Calculations. Sensors 2013, 13, 9321–9343. [Google Scholar] [CrossRef]
- Nwanna, O. Validation of an Accelerometry Based Method of Human Gait Analysis. Ph.D. Thesis, Cleveland State University, Cleveland, OH, USA, 2014. [Google Scholar]
- Ju, H.; Lee, M.S.; Park, S.Y.; Song, J.W.; Park, C.G. A pedestrian dead-reckoning system that considers the heel-strike and toe-off phases when using a foot-mounted IMU. Meas. Sci. Technol. 2016, 27, 015702. [Google Scholar] [CrossRef]
- Park, S.Y.; Ju, H.; Park, C.G. Stance Phase Detection of Multiple Actions for Military Drill Using Foot-mounted IMU. In Proceedings of the International Conference on Indoor Positioning and Indoor Navigation, Alcala de Henares, Spain, 4–7 October 2016; p. 4. [Google Scholar]
- Zhao, H.; Wang, Z.; Qiu, S.; Shen, Y.; Wang, J. IMU-based gait analysis for rehabilitation assessment of patients with gait disorders. In Proceedings of the 2017 4th International Conference on Systems and Informatics (ICSAI), Hangzhou, China, 11–13 November 2017; pp. 622–626. [Google Scholar] [CrossRef]
- Mei, C.; Gao, F.; Li, Y. A Determination Method for Gait Event Based on Acceleration Sensors. Sensors 2019, 19, 5499. [Google Scholar] [CrossRef] [PubMed]
- Zhen, T.; Yan, L.; Yuan, P. Walking Gait Phase Detection Based on Acceleration Signals Using LSTM-DNN Algorithm. Algorithms 2019, 12, 253. [Google Scholar] [CrossRef]
- Perez-Ibarra, J.C.; Siqueira, A.A.G.; Krebs, H.I. Real-Time Identification of Gait Events in Impaired Subjects Using a Single-IMU Foot-Mounted Device. IEEE Sens. J. 2020, 20, 2616–2624. [Google Scholar] [CrossRef]
- Liu, X.; Li, N.; Xu, G.; Zhang, Y. A Novel Robust Step Detection Algorithm for Foot-Mounted IMU. IEEE Sens. J. 2021, 21, 5331–5339. [Google Scholar] [CrossRef]
- Wouda, F.J.; Jaspar, S.L.J.O.; Harlaar, J.; van Beijnum, B.J.F.; Veltink, P.H. Foot progression angle estimation using a single foot-worn inertial sensor. J. Neuroeng. Rehabil. 2021, 18, 37. [Google Scholar] [CrossRef]
- Grimmer, M.; Schmidt, K.; Duarte, J.E.; Neuner, L.; Koginov, G.; Riener, R. Stance and Swing Detection Based on the Angular Velocity of Lower Limb Segments During Walking. Front. Neurorobotics 2019, 13, 57. [Google Scholar] [CrossRef]
- Romijnders, R.; Warmerdam, E.; Hansen, C.; Welzel, J.; Schmidt, G.; Maetzler, W. Validation of IMU-based gait event detection during curved walking and turning in older adults and Parkinson’s Disease patients. J. Neuroeng. Rehabil. 2021, 18, 28. [Google Scholar] [CrossRef]
- Torrealba, R.; Castellano, J.; Fernández-López, G.; Grieco, J. Characterisation of gait cycle from accelerometer data. Electron. Lett. 2007, 43, 1066. [Google Scholar] [CrossRef]
- Wentink, E.; Beijen, S.; Hermens, H.; Rietman, J.; Veltink, P. Intention detection of gait initiation using EMG and kinematic data. Gait Posture 2013, 37, 223–228. [Google Scholar] [CrossRef]
- Ryu, J.; Lee, B.H.; Maeng, J.; Kim, D.H. sEMG-signal and IMU sensor-based gait sub-phase detection and prediction using a user-adaptive classifier. Med. Eng. Phys. 2019, 69, 50–57. [Google Scholar] [CrossRef] [PubMed]
- Yang, C.C.; Hsu, Y.L.; Shih, K.S.; Lu, J.M. Real-Time Gait Cycle Parameter Recognition Using a Wearable Accelerometry System. Sensors 2011, 11, 7314–7326. [Google Scholar] [CrossRef]
- Hwang, T.H.; Reh, J.; Effenberg, A.; Blume, H. Real-time gait event detection using a single head-worn inertial measurement unit. In Proceedings of the 2016 IEEE 6th International Conference on Consumer Electronics-Berlin (ICCE-Berlin), Berlin, Germany, 5–7 September 2016; pp. 28–32. [Google Scholar] [CrossRef]
- Hwang, T.H.; Reh, J.; Effenberg, A.O.; Blume, H. Validation of Real Time Gait Analysis Using a Single Head-Worn IMU. In EKC 2019 Conference Proceedings; Park, J.M., Whang, D.R., Eds.; Springer: Singapore, 2021; pp. 87–97. [Google Scholar] [CrossRef]
- Schepers, M.; Giuberti, M.; Bellusci, G. Xsens MVN: Consistent Tracking of Human Motion Using Inertial Sensing. 2018; unpublished. [Google Scholar] [CrossRef]
- Suwansawang, S. Analysis of Human Locomotion Using Analytic Wavelets Applied to Electromyographic Data from Healthy Controls and Parkinson’s Patients. Ph.D. Thesis, University of York, York, UK, 2018. [Google Scholar]
- Morbidoni, C.; Cucchiarelli, A.; Fioretti, S.; Di Nardo, F. A Deep Learning Approach to EMG-Based Classification of Gait Phases during Level Ground Walking. Electronics 2019, 8, 894. [Google Scholar] [CrossRef]
- Brantley, J.A.; Luu, T.P.; Nakagome, S.; Zhu, F.; Contreras-Vidal, J.L. Full body mobile brain-body imaging data during unconstrained locomotion on stairs, ramps, and level ground. Sci. Data 2018, 5, 180133. [Google Scholar] [CrossRef] [PubMed]
- Bächlin, M.; Plotnik, M.; Roggen, D.; Maidan, I.; Hausdorff, J.M.; Giladi, N.; Tröster, G. Wearable Assistant for Parkinson’s Disease Patients with the Freezing of Gait Symptom. Trans. Inf. Technol. Biomed. 2010, 14, 436–446. [Google Scholar] [CrossRef]
- XSens. MVN User Manual; XSens: Enschede, The Netherlands, 2021. [Google Scholar]
- MATLAB, Version 9.8.0 (R2020a); The MathWorks Inc.: Natick, MA, USA, 2020.
- Marmelat, V.; Duncan, A.; Meltz, S. Effect of sampling frequency on fractal fluctuations during treadmill walking. PLoS ONE 2019, 14, e0218908. [Google Scholar] [CrossRef] [PubMed]
- Zhou, L.; Fischer, E.; Tunca, C.; Brahms, C.M.; Ersoy, C.; Granacher, U.; Arnrich, B. How we found our imu: Guidelines to IMU selection and a comparison of seven Imus for Pervasive Healthcare Applications. Sensors 2020, 20, 4090. [Google Scholar] [CrossRef] [PubMed]
- Clancy, E.; Morin, E.; Merletti, R. Sampling, noise-reduction and amplitude estimation issues in surface electromyography. J. Electromyogr. Kinesiol. 2002, 12, 1–16. [Google Scholar] [CrossRef]
- Dua, D.; Graff, C. UCI Machine Learning Repository; School of Information and Computer Science, University of California: Irvine, CA, USA, 2019; Available online: http://archive.ics.uci.edu/ml (accessed on 18 April 2023).
Sensor Type | Advantages | Disadvantages |
---|---|---|
Optical Systems | Full-body kinematics. High marker detection precision and accuracy. Accurate estimation of joint position, angle, velocity and acceleration. Synchronisation solutions with other gait assessment hardware (EMG kits). | High cost and maintenance. High performance requirements. Restricted experimental environments. Multiple camera synchronisation and calibration. Setup and collection learning curve. Imposes long donning and doffing times of retroreflective markers, as well as initial setup per subject. |
FSRs and IPSs | Ideal for detection of key gait events as heel-strike and toe-off. Low-cost solution, with lower number of sensors required for event detection. Typical benchmark for other gait assessment technologies | Low durability, sensor wearing out. Depending on sensor application site, may impact the subject’s walking style. Uneven terrain and uncharacteristic gait behaviours may lead to accidental noise event recordings. |
IMUs | Simple configuration and multitude of solutions to interface with other hardware (EMG kits). Low-cost and high availability. Applicability to diverse gait environments, especially when using wearable solutions. Easy and quick donning and doffing of the sensors to and across subjects. | Lower accuracy. Wearable solutions are battery dependent. High susceptibility to motion artefact noises, depending on application site and surface (directly on the skin or over clothing). Position estimation requires multiple synchronised sensors. |
Surface EMG | Valuable physiological information of muscle activity patterns involved in gait. May be used to identify personal muscle recruitment schemes used per individual. | Extremely high inter-subject variability. Hard to characterise gait by itself and requires significant subject by subject adjustments. |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Castro Aguiar, R.; Sam Jeeva Raj, E.J.; Chakrabarty, S. Simplified Markerless Stride Detection Pipeline (sMaSDP) for Surface EMG Segmentation. Sensors 2023, 23, 4340. https://doi.org/10.3390/s23094340
Castro Aguiar R, Sam Jeeva Raj EJ, Chakrabarty S. Simplified Markerless Stride Detection Pipeline (sMaSDP) for Surface EMG Segmentation. Sensors. 2023; 23(9):4340. https://doi.org/10.3390/s23094340
Chicago/Turabian StyleCastro Aguiar, Rafael, Edward Jero Sam Jeeva Raj, and Samit Chakrabarty. 2023. "Simplified Markerless Stride Detection Pipeline (sMaSDP) for Surface EMG Segmentation" Sensors 23, no. 9: 4340. https://doi.org/10.3390/s23094340