EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Proof-of-Concept End-to-End EEG-BCI Pipeline for Grasp-and-Release Decoding and Low-Cost Tendon-Driven Actuation

Krishnakumar Sankar, Kanimozhi Gowtheman, Keerthivasan Sundararajan, Naresh Moorthi

Smart Wearable Technology · 30.10.2026

DOI: 10.47852/bonviewswt620210799 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Smart Wearable Technology, online erschienen am 30.10.2026. Verfasst von Krishnakumar Sankar und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Stroke is a leading cause of long-term motor disability, inspiring research in the design of assistive and rehabilitation technologies that work in situations with limited residual muscle activity. Brain–computer interfaces (BCIs) offer a promising non-muscular pathway to translate brain activity into BCI commands. This paper presents the proof-of-concept end-to-end pipeline for the use of binary grasp-and-release Electroencephalography-Brain Computer Interface (EEG-BCI) decoding in conjunction with low-cost tendon-driven robotic hand actuation. The classification part of our pipeline has been trained on the publicly available WAY EEG Grasp-and-Lift dataset. It includes EEG recordings of performed grasp-and-lift movements of healthy participants, instead of motor-imagery movements. Eight EEG channels have been filtered, common average referenced, and normalized and then classified in a subject-independent manner using the EEGNet network. The proposed EEGNet achieved a trial-level accuracy of 74.74%, outperforming the baseline (60.38%) FBCSP + SVM classifier. In terms of hardware, the acknowledgment success rate was 98.18%, with 54 out of 55 transmitted commands recieving an ACK response. The executed-movement EEG data have been used only as a demonstration of engineering proof-of-concept for our decoding and robotic control pipeline. Motor-imagery EEG decoding, patients’ evaluation, and clinical outcomes have not been considered in this study. Received: 10 June 2026 | Revised: 14 September 2026 | Accepted: 9 October 2026 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement The data that support the findings of this study are openly available at https://www.kaggle.com/c/grasp-and-lift-eeg-detection/data. Author Contribution Statement Krishnakumar Sankar: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Kanimozhi Gowtheman: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft. Keerthivasan Sundararajan: Conceptualization, Writing – original draft. Naresh Moorthi: Conceptualization, Writing – original draft.

Zitieren

APA

Sankar, K., Gowtheman, K., Sundararajan, K., & Moorthi, N. (2026). Proof-of-Concept End-to-End EEG-BCI Pipeline for Grasp-and-Release Decoding and Low-Cost Tendon-Driven Actuation. Smart Wearable Technology. https://doi.org/10.47852/bonviewswt620210799

Vancouver

Sankar K, Gowtheman K, Sundararajan K, Moorthi N. Proof-of-Concept End-to-End EEG-BCI Pipeline for Grasp-and-Release Decoding and Low-Cost Tendon-Driven Actuation. Smart Wearable Technology. 2026. doi:10.47852/bonviewswt620210799

BibTeX
@article{Sankar2026,
  author = {Sankar, Krishnakumar and Gowtheman, Kanimozhi and Sundararajan, Keerthivasan and Moorthi, Naresh},
  title = {Proof-of-Concept End-to-End EEG-BCI Pipeline for Grasp-and-Release Decoding and Low-Cost Tendon-Driven Actuation},
  journal = {Smart Wearable Technology},
  year = {2026},
  doi = {10.47852/bonviewswt620210799}
}
Kostenlos starten →