Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Wavelet Analysis of Noninvasive EEG Signals Discriminates Complex and Natural Grasp Types

Ali Rabiee, Sima Ghafoori, Anna Cetera, Reza Abiri

TL;DR

This study addresses decoding of complex hand grasps from EEG to advance dexterous neuroprosthetics and BCIs. It combines complex Morlet wavelet-based time-frequency analysis with 8-channel EEG and multiple classifiers (SVM, RF, XGBoost, LDA), validated by 5-fold cross-validation and permutation feature importance. Results show high accuracies, with multiclass classification at 85.16% and binary distinctions exceeding 95% for No-Movement vs grasp types, identifying motor cortex alpha/beta features (notably at C3/C4) as key discriminators around 300 ms post-movement. The work demonstrates the potential of wavelet-based EEG features for real-time neuroprosthetic interfaces, while noting the need for larger cohorts and real-time system integration in future studies.

Abstract

This research aims to decode hand grasps from Electroencephalograms (EEGs) for dexterous neuroprosthetic development and Brain-Computer Interface (BCI) applications, especially for patients with motor disorders. Particularly, it focuses on distinguishing two complex natural power and precision grasps in addition to a neutral condition as a no-movement condition using a new EEG-based BCI platform and wavelet signal processing. Wavelet analysis involved generating time-frequency and topographic maps from wavelet power coefficients. Then, by using machine learning techniques with novel wavelet features, we achieved high average accuracies: 85.16% for multiclass, 95.37% for No-Movement vs Power, 95.40% for No-Movement vs Precision, and 88.07% for Power vs Precision, demonstrating the effectiveness of these features in EEG-based grasp differentiation. In contrast to previous studies, a critical part of our study was permutation feature importance analysis, which highlighted key features for grasp classification. It revealed that the most crucial brain activities during grasping occur in the motor cortex, within the alpha and beta frequency bands. These insights demonstrate the potential of wavelet features in real-time neuroprosthetic technology and BCI applications.

Wavelet Analysis of Noninvasive EEG Signals Discriminates Complex and Natural Grasp Types

TL;DR

This study addresses decoding of complex hand grasps from EEG to advance dexterous neuroprosthetics and BCIs. It combines complex Morlet wavelet-based time-frequency analysis with 8-channel EEG and multiple classifiers (SVM, RF, XGBoost, LDA), validated by 5-fold cross-validation and permutation feature importance. Results show high accuracies, with multiclass classification at 85.16% and binary distinctions exceeding 95% for No-Movement vs grasp types, identifying motor cortex alpha/beta features (notably at C3/C4) as key discriminators around 300 ms post-movement. The work demonstrates the potential of wavelet-based EEG features for real-time neuroprosthetic interfaces, while noting the need for larger cohorts and real-time system integration in future studies.

Abstract

This research aims to decode hand grasps from Electroencephalograms (EEGs) for dexterous neuroprosthetic development and Brain-Computer Interface (BCI) applications, especially for patients with motor disorders. Particularly, it focuses on distinguishing two complex natural power and precision grasps in addition to a neutral condition as a no-movement condition using a new EEG-based BCI platform and wavelet signal processing. Wavelet analysis involved generating time-frequency and topographic maps from wavelet power coefficients. Then, by using machine learning techniques with novel wavelet features, we achieved high average accuracies: 85.16% for multiclass, 95.37% for No-Movement vs Power, 95.40% for No-Movement vs Precision, and 88.07% for Power vs Precision, demonstrating the effectiveness of these features in EEG-based grasp differentiation. In contrast to previous studies, a critical part of our study was permutation feature importance analysis, which highlighted key features for grasp classification. It revealed that the most crucial brain activities during grasping occur in the motor cortex, within the alpha and beta frequency bands. These insights demonstrate the potential of wavelet features in real-time neuroprosthetic technology and BCI applications.
Paper Structure (12 sections, 3 figures, 2 tables)

This paper contains 12 sections, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methods
  3. Participants
  4. Experimental Setup and Paradigm
  5. Data Recording and Preprocessing
  6. Wavelet Analysis
  7. Classification
  8. Results
  9. Wavelet Analysis
  10. Classification
  11. Discussion
  12. Conclusion

Figures (3)

  • Figure 1: Experimental setup and paradigm for reach-and-grasp tasks. (a) Platform components, (b) an illustration of a 128-channel EEG sensor layout with the 8 electrodes (in red) that are used in the study, and (c) The timeline of audio Cues and actions in the experimental paradigm.
  • Figure 2: The power of wavelet coefficients in EEG during hand-grasping tasks, with time-frequency plots for different conditions, and corresponding topographic maps focusing on 300ms post-movement across various frequencies.
  • Figure 3: (a) Boxplot illustrating the permutation feature importance scores for the top 10 EEG features in the classification of hand-grasping actions. (b) Topographic maps showing the spatial distribution of feature importance across the brain at different frequency bands.