Deep Learning Models Optimization for Gait Phase Identification from EMG Data During Exoskeleton-Assisted Walking
Biomimetics · 13.09.2025 · 10 (9) · 617
DOI: 10.3390/biomimetics10090617 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biomimetics, online erschienen am 13.09.2025. Verfasst von Roberto Soldi und 8 weiteren Autor:innen. Beteiligte Einrichtung: Laboratory of Bioengineering, Department of Electrical, Computer and Biomedical Engineering, University of Pavia, 27100 Pavia, Italy. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 75 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministero dell'università e della ricerca, Italian Ministry of University and Research.
Abstract (Originalsprache)
Exoskeletons are a fast-growing technology that enables multiple use-cases in clinical scenarios. They can be useful tools for the rehabilitation of patients with motor dysfunctions caused by neurological conditions, aging or trauma. Assistive exoskeletons modulate the torque exerted by the electrical motors moving their joints to allow the patients wearing them to achieve an intended movement, such as gait, correctly. Their effectiveness, therefore, requires accurate online control of such torques to complement those generated by the patient. Hereby we explored Deep Learning (DL) models to generate an online prediction of the gait phase, i.e., stance or swing, during assisted walking with a lower-limb exoskeleton based on surface electromyography (sEMG) data. We leveraged the lead of muscular activation with respect to the movement of the limbs to adjust the labeling based on joints kinematics. The cross-subject design allowed to generalize over subjects not considered for training A hyperparameter optimization algorithm was also implemented to further explore the capabilities of DL models of a reduced size. We simulated a use case scenario to assess whether online implementation of the proposed technique is feasible. We also proposed a new metric called trade-of score (TOS) for evaluating the cost-performance compromise of the optimized models which lead to identifying a DL model capable of classifying gait phases with an accuracy of about 95% while significantly reducing the number of parameters compared to the full architecture. Its mean computational time of less than 10 ms offers the opportunity for accurate, online exoskeleton control based on sEMG data.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Laboratory of Bioengineering, Department of Electrical, Computer and Biomedical Engineering, University of Pavia, 27100 Pavia, Italy.
- Department of Research and Development, LUNEX International University of Health, Exercise and Sports, Avenue du Parc des Sports, 50, 4671 Differdange, Luxembourg.
- Luxembourg Health & Sport Sciences Research Institute ASBL, Avenue du Parc des Sports, 50, 4671 Differdange, Luxembourg.
ORCID-Profile: Roberto Soldi, Bruna Maria Vittoria Guerra, Leo Russo, Renato Baptista, Alessandro Marco De Nunzio, Stefano Ramat
Förderung
- Ministero dell'università e della ricerca
- Italian Ministry of University and Research
Zitieren
APA
Soldi, R., Guerra, B. M. V., Sozzi, S., Russo, L., Pizzocaro, S., Baptista, R., De Nunzio, A. M., Schmid, M., & Ramat, S. (2025). Deep Learning Models Optimization for Gait Phase Identification from EMG Data During Exoskeleton-Assisted Walking. Biomimetics, 10(9), 617. https://doi.org/10.3390/biomimetics10090617
Vancouver
Soldi R, Guerra BMV, Sozzi S, Russo L, Pizzocaro S, Baptista R, et al. Deep Learning Models Optimization for Gait Phase Identification from EMG Data During Exoskeleton-Assisted Walking. Biomimetics. 2025;10(9):617. doi:10.3390/biomimetics10090617
BibTeX
@article{Soldi2025,
author = {Soldi, Roberto and Guerra, Bruna Maria Vittoria and Sozzi, Stefania and Russo, Leo and Pizzocaro, Serena and Baptista, Renato and De Nunzio, Alessandro Marco and Schmid, Micaela and Ramat, Stefano},
title = {Deep Learning Models Optimization for Gait Phase Identification from EMG Data During Exoskeleton-Assisted Walking},
journal = {Biomimetics},
year = {2025},
volume = {10},
number = {9},
pages = {617},
doi = {10.3390/biomimetics10090617}
}
EUVIMED