AccScience Publishing / EJMO / Online First / DOI: 10.36922/EJMO025410433
Submit to EJMO
Cite this article
3
Download
31
Views
Journal Browser
Volume | Year
Issue
Search
Forthcoming Issue
Current Issue
View All
News and Announcements
View All
ORIGINAL RESEARCH ARTICLE

A hybrid convolutional neural network–bidirectional long short-term memory framework for multimodal gait phase recognition

Aihui Wang1† ,  Xiang Zhang1† ,  Zhaowei Sun2* ,  Hengyi Li1 ,  Jinkang Dong1 ,  Shengda Gao3
Show Less
1 Department of Automation, School of Automation and Electrical Engineering, Zhongyuan University of Technology, Zhengzhou, Henan, China
2 Henan Huadong Industrial Control Technology Company Ltd, Zhengzhou, Henan, China
3 Department of Electronic and Computer Engineering, Graduate School of Science and Engineering, Ritsumeikan University, Kusatsu, Shiga, Japan
†These authors contributed equally to this work.
EJMO, 025410433 https://doi.org/10.36922/EJMO025410433
Received: 11 October 2025 | Revised: 28 December 2025 | Accepted: 27 February 2026 | Published online: 15 April 2026
© 2026 by the Author(s). This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution -Noncommercial 4.0 International License (CC-by the license) ( https://creativecommons.org/licenses/by-nc/4.0/ )
Download PDF
XML
Cite
Abstract

Introduction: Surface electromyography (sEMG) signals are a key modality for gait phase recognition, as they reflect neuromuscular activation and capture gait dynamics essential for natural control in rehabilitation robotics and human–machine systems. However, reliable phase recognition under complex locomotion conditions remains challenging due to high inter-subject variability, ambiguous phase boundaries, and increased biomechanical diversity across gait types.

Objective: To address this issue, this paper proposes a hybrid deep learning framework that leverages the complementary strengths of convolutional neural networks (CNNs) for local time-series feature extraction and bidirectional long short-term memory networks (BiLSTMs) for modeling bidirectional long-range dependencies.

Methods: A dataset was collected from 11 healthy subjects performing five representative locomotion scenarios, with gait phases annotated using synchronized motion capture and plantar pressure signals.

Results: The proposed method achieved an average classification accuracy of 95.81%, with performance variation across gait types within 4%. The corresponding averaged precision, recall, and F1-score were 94.35%, 95.66%, and 94.95%, respectively, indicating stable and well-balanced recognition performance across multiple locomotion scenarios. Notably, the framework maintained robust performance in biomechanically challenging conditions such as slope ascent and stair ambulation, demonstrating its effectiveness under complex gait dynamics.

Conclusion: Overall, these results demonstrate that the proposed framework provides a stable and generalizable solution for sEMG-based gait phase recognition in complex locomotion environments.

Keywords
Gait phase recognition
Surface electromyography
Deep learning
Funding
This research was funded by the Henan Province Key Research and Development Project (Grants No. 241111312000), Henan Province Key International Science and Technology Cooperation Project (Grants No. 251111520400), the Key R&D Project of Henan Province (Grants No. 251111220900), and the Henan Province Key Technologies Research Development project (Grants No. 252102211106, 252102320281).
Conflict of interest
The authors declare they have no competing interests.
References
  1. Roberts M, Mongeon D, Prince F. Biomechanical parameters for gait analysis: a systematic review of healthy human gait. Phys Ther Rehabil. 2017;4(1):6. doi:10.7243/2055-2386-4-6

 

  1. Iseki C, Hayasaka T, Yanagawa H, et al. Artificial intelligence distinguishes pathological gait: the analysis of markerless motion capture gait data acquired by an iOS application (TDPT-GT). Sensors (Basel). 2023;23(13):6217. doi: 10.3390/s23136217

 

  1. Jing Y, Qin P, Fan X, et al. Deep learning–assisted gait parameter assessment for neurodegenerative diseases: model development and validation. J Med Internet Res. 2023;25:e46427. doi: 10.2196/46427

 

  1. Kim Y, Kim J, Moon J, et al. Deep learning-based recognition of locomotion mode, phase, and phase progression using inertial measurement units. J Bionic Eng. 2025;22(4):1804- 1818. doi: 10.1007/s42235-025-00723-7

 

  1. Di Gregorio R, Vocenas L. Identification of gait-cycle phases for prosthesis control. Biomimetics (Basel). 2021;6(2):22. doi: 10.3390/biomimetics6020022

 

  1. Vu HTT, Lee H, Park S, et al. A review of gait phase detection algorithms for lower-limb prostheses and exoskeletons. Sensors (Basel). 2020;20(14):3972. doi: 10.3390/s20143972

 

  1. Hulleck AA, Menoth Mohan D, Abdallah N, El Rich M, Khalaf K. Present and future of gait assessment in clinical practice: toward the application of novel trends and technologies. Front Med Technol. 2022;4:901331. doi: 10.3389/fmedt.2022.901331

 

  1. Brambilla C, Beltrame G, Marino G, et al. Biomechanical analysis of human gait when changing velocity and carried loads: simulation study with OpenSim. Biology (Basel). 2024;13(5):321. doi: 10.3390/biology13050321

 

  1. Perry J, Burnfield JM. The gait cycle and its variations with disease and injury. J Bodyw Mov Ther. 2020;24(2):101-113. doi: 10.1016/j.mporth.2020.03.009

 

  1. Aqueveque P, Germany E, Osorio R, et al. Gait segmentation method using a plantar pressure measurement system with custom-made capacitive sensors. Sensors (Basel). 2020;20(3):656. doi: 10.3390/s20030656

 

  1. Wu J, Becsek B, Schaer A, et al. Real-time gait phase detection on wearable devices for real-world free-living gait. IEEE J Biomed Health Inform. 2023;27(3):1295-1306. doi: 10.1109/JBHI.2022.3228329

 

  1. Choi S, Ko C, Kong K. Walking-speed-adaptive gait phase estimation for wearable robots. Sensors (Basel). 2023;23(19):8276. doi: 10.3390/s23198276

 

  1. Xu S, Fang J, Hu X, et al. Emotion recognition from gait analyses: current research and future directions. IEEE Trans Comput Soc Syst. 2024;11(1):363-377. doi: 10.1109/TCSS.2022.3223251

 

  1. Benedetti MG, Agostini V, Knaflitz M, Bonato P. Muscle Activation Patterns During Level Walking and Stair Ambulation. In: Applications of EMG in Clinical and Sports Medicine. InTech; 2012. doi: 10.5772/25792

 

  1. Escamilla-Martínez E, Martínez-Nova A, Gómez-Martín B, et al. The effect of moderate running on foot posture index and plantar pressure distribution in male recreational runners. J Am Podiatr Med Assoc. 2013;103(2):121-125. doi: 10.7547/1030121

 

  1. Crea S, Donati M, De Rossi S, Oddo C, Vitiello N. A wireless flexible sensorized insole for gait analysis. Sensors. 2014;14(1):1073-1093. doi: 10.3390/s140101073

 

  1. Sawicki A, Saeed K. Application of LSTM networks for human gait-based identification. In: Advances in Intelligent Systems and Computing. Springer; 2021:402-412. doi: 10.1007/978-3-030-76773-0_39

 

  1. Guimarães V, Sousa I, Correia MV. A deep learning approach for foot trajectory estimation in gait analysis using inertial sensors. Sensors (Basel). 2021;21(22):7517. doi: 10.3390/s21227517

 

  1. Han X, Guffanti D, Brunete A. A comprehensive review of vision-based sensor systems for human gait analysis. Sensors (Basel). 2025;25(2):498. doi: 10.3390/s25020498

 

  1. Yang S, Li Q. Inertial sensor-based methods in walking speed estimation: a systematic review. Sensors (Basel). 2012;12(5):6102-6116. doi: 10.3390/s120506102

 

  1. Kyeong S, Shin W, Yang M, et al. Recognition of walking environments and gait period by surface electromyography. Front Inf Technol Electron Eng. 2019;20(3):342-352. doi: 10.1631/FITEE.1800601

 

  1. Zhu J, Yi C, Wei B, et al. The muscle fatigue’s effects on the sEMG-based gait phase classification: an experimental study and a novel training strategy. Appl Sci (Basel). 2021;11(9):3821. doi: 10.3390/app11093821

 

  1. Shi X, Zhang X, Qin P, et al. Gait phase recognition in multi- task scenarios based on sEMG signals. Biosensors (Basel). 2025;15(5):305. doi: 10.3390/bios15050305

 

  1. Wei B, Ding Z, Yi C, et al. A novel sEMG-based gait phase- kinematics-coupled predictor and its interaction with exoskeletons. Front Neurorobot. 2021;15:704226. doi: 10.3389/fnbot.2021.704226

 

  1. Song Q, Ma X, Liu Y. Continuous online prediction of lower limb joints angles based on sEMG signals by deep learning approach. Comput Biol Med. 2023;163:107124. doi: 10.1016/j.compbiomed.2023.107124

 

  1. Carvalho CR, Fernández JM, del-Ama AJ, et al. Review of electromyography onset detection methods for real- time control of robotic exoskeletons. J Neuroeng Rehabil. 2023;20:141. doi: 10.1186/s12984-023-01268-8

 

  1. Agostini V, Ghislieri M, Rosati S, et al. Surface electromyography applied to gait analysis: how to improve its impact in clinics? Front Neurol. 2020;11:994. doi: 10.3389/fneur.2020.00994

 

  1. Phinyomark A, Khushaba RN, Scheme E. Feature extraction and selection for myoelectric control based on wearable EMG sensors. Sensors (Basel). 2018;18(5):1615. doi: 10.3390/s18051615

 

  1. Mesin L. Crosstalk in surface electromyogram: literature review and some insights. Phys Eng Sci Med. 2020;43(2):481- 492. doi: 10.1007/s13246-020-00868-1

 

  1. Germer CM, Farina D, Elias LA, et al. Surface EMG cross talk quantified at the motor unit population level for muscles of the hand, thigh, and calf. J Appl Physiol. 2021;131(2):808- 820. doi: 10.1152/japplphysiol.01041.2020

 

  1. McManus L, De Vito G, Lowery MM. Analysis and biophysics of surface EMG for physiotherapists and kinesiologists: toward  a  common  language  with rehabilitation engineers. Front Neurol. 2020;11:576729. doi: 10.3389/fneur.2020.576729

 

  1. Sun J, Liu G, Sun Y, et al. Application of surface electromyography in exercise fatigue: a review. Front Syst Neurosci. 2022;16:893275. doi: 10.3389/fnsys.2022.893275

 

  1. Qassim HM, Hasan WZW, Ramli HR, et al. Proposed fatigue index for the objective detection of muscle fatigue using surface electromyography and a double-step binary classifier. Sensors (Basel). 2022;22(5):1900. doi: 10.3390/s22051900

 

  1. Zhou H, Yang D, Li Z, et al. Locomotion mode recognition for walking on three terrains based on sEMG of lower limb and back muscles. Sensors (Basel). 2021;21(9):2933. doi: 10.3390/s21092933

 

  1. Kim P, Lee J, Shin CS. Classification of walking environments using deep learning approach based on surface EMG sensors only. Sensors (Basel). 2021;21(12):4204. doi: 10.3390/s21124204

 

  1. Matsuda T, Miyamori T, Fujino Y, et al. Reliability and validity of muscle activity analysis using wearable electromyographs. J Phys Ther Sci. 2024;36(3):123-127. doi: 10.1589/jpts.36.123

 

  1. Di Nardo F, Morbidoni C, Fioretti S. Surface electromyographic signals collected during long-lasting ground walking of young able-bodied subjects. PhysioNet. Posted online 2024. doi: 10.13026/GRDJ-QX06

 

  1. Di Nardo F, Strazza A, Mengarelli A, et al. Surface EMG patterns for quantification of thigh muscle co-contraction in school-age children: normative data during walking. Gait Posture. 2018;61:25-33. doi: 10.1016/j.gaitpost.2017.12.025

 

  1. Gkrekidis A, Giarmatzis G, Menychtas D, et al. Clinical validation of estimated muscle activations during phases of elderly gait. Biomechanics. 2023;3(4):552-560. doi: 10.3390/biomechanics3040044

 

  1. Costa-García Á, Iáñez E, Sonoo M, et al. Segmentation and averaging of sEMG muscle activations prior to synergy extraction. IEEE Robot Autom Lett. 2020;5(2):3106-3112. doi: 10.1109/LRA.2020.2975729

 

  1. Tokas P, Semwal VB, Jain S, et al. Revolutionizing gait analysis with blockchain technology: enhancing the privacy and security. In: Sustainable Security Practices Using Blockchain, Quantum and Post-Quantum Technologies for Real Time Applications. Springer; 2024:239-254. doi: 10.1007/978-981-97-0088-2_12

 

  1. Reddy MVM, Hariharan S, Kaliyaperuma G, Pandian PS, Kumaravel S, Karthick PA. Recognition of gait phases from time-domain descriptors of surface EMG signals and dynamic convolutional neural networks. In: 2025 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). IEEE; 2025:1-4. doi: 10.1109/embc58623.2025.11251734

 

  1. Wei P, Zhang J, Wang B, et al. Surface electromyography and electroencephalogram-based gait phase recognition and correlations between cortical and locomotor muscle in the seven gait phases. Front Neurosci. 2021;15:607905. doi: 10.3389/fnins.2021.607905

 

  1. Cai S, Chen D, Fan B, et al. Gait phases recognition based on lower limb sEMG signals using LDA-PSO-LSTM algorithm. Biomed Signal Process Control. 2023;80:104272. doi: 10.1016/j.bspc.2022.104272
Share
Back to top
Eurasian Journal of Medicine and Oncology, Electronic ISSN: 2587-196X Print ISSN: 2587-2400, Published by AccScience Publishing
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here
Accept