Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

FingerNet: EEG Decoding of A Fine Motor Imagery with Finger-tapping Task Based on A Deep Neural Network

Young-Min Go, Seong-Hyun Yu, Hyeong-Yeong Park, Minji Lee, Ji-Hoon Jeong

TL;DR

FingerNet is introduced, a specialized network for fine MI classification, departing from conventional gross MI studies, and the weighted Cross Entropy approach employed to address such biased predictions appears to have broader applicability and relevance across various domains involving multi-class classification tasks.

Abstract

Brain-computer interface (BCI) technology facilitates communication between the human brain and computers, primarily utilizing electroencephalography (EEG) signals to discern human intentions. Although EEG-based BCI systems have been developed for paralysis individuals, ongoing studies explore systems for speech imagery and motor imagery (MI). This study introduces FingerNet, a specialized network for fine MI classification, departing from conventional gross MI studies. The proposed FingerNet could extract spatial and temporal features from EEG signals, improving classification accuracy within the same hand. The experimental results demonstrated that performance showed significantly higher accuracy in classifying five finger-tapping tasks, encompassing thumb, index, middle, ring, and little finger movements. FingerNet demonstrated dominant performance compared to the conventional baseline models, EEGNet and DeepConvNet. The average accuracy for FingerNet was 0.3049, whereas EEGNet and DeepConvNet exhibited lower accuracies of 0.2196 and 0.2533, respectively. Statistical validation also demonstrates the predominance of FingerNet over baseline networks. For biased predictions, particularly for thumb and index classes, we led to the implementation of weighted cross-entropy and also adapted the weighted cross-entropy, a method conventionally employed to mitigate class imbalance. The proposed FingerNet involves optimizing network structure, improving performance, and exploring applications beyond fine MI. Moreover, the weighted Cross Entropy approach employed to address such biased predictions appears to have broader applicability and relevance across various domains involving multi-class classification tasks. We believe that effective execution of motor imagery can be achieved not only for fine MI, but also for local muscle MI

FingerNet: EEG Decoding of A Fine Motor Imagery with Finger-tapping Task Based on A Deep Neural Network

TL;DR

FingerNet is introduced, a specialized network for fine MI classification, departing from conventional gross MI studies, and the weighted Cross Entropy approach employed to address such biased predictions appears to have broader applicability and relevance across various domains involving multi-class classification tasks.

Abstract

Brain-computer interface (BCI) technology facilitates communication between the human brain and computers, primarily utilizing electroencephalography (EEG) signals to discern human intentions. Although EEG-based BCI systems have been developed for paralysis individuals, ongoing studies explore systems for speech imagery and motor imagery (MI). This study introduces FingerNet, a specialized network for fine MI classification, departing from conventional gross MI studies. The proposed FingerNet could extract spatial and temporal features from EEG signals, improving classification accuracy within the same hand. The experimental results demonstrated that performance showed significantly higher accuracy in classifying five finger-tapping tasks, encompassing thumb, index, middle, ring, and little finger movements. FingerNet demonstrated dominant performance compared to the conventional baseline models, EEGNet and DeepConvNet. The average accuracy for FingerNet was 0.3049, whereas EEGNet and DeepConvNet exhibited lower accuracies of 0.2196 and 0.2533, respectively. Statistical validation also demonstrates the predominance of FingerNet over baseline networks. For biased predictions, particularly for thumb and index classes, we led to the implementation of weighted cross-entropy and also adapted the weighted cross-entropy, a method conventionally employed to mitigate class imbalance. The proposed FingerNet involves optimizing network structure, improving performance, and exploring applications beyond fine MI. Moreover, the weighted Cross Entropy approach employed to address such biased predictions appears to have broader applicability and relevance across various domains involving multi-class classification tasks. We believe that effective execution of motor imagery can be achieved not only for fine MI, but also for local muscle MI
Paper Structure (8 sections, 3 equations, 5 figures, 2 tables)

This paper contains 8 sections, 3 equations, 5 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Materials and Methods
  3. Experimental Protocols
  4. EEG Preprocessing
  5. Model Structure
  6. Loss Function
  7. Results and Discussion
  8. Conclusion and Future Works

Figures (5)

  • Figure 1: Experimental protocol comprises the phases of rest (3s), fixation (2s), instruction (2s), task (4s). The session incorporates task for all fingers of the right hand with 25 tasks per finger and a total of 250 trials.
  • Figure 2: Setting for acquiring EEG data during the experiment.
  • Figure 3: Performance comparison of weighted cross-entropy using each set of weights.
  • Figure 4: Performance comparison of weighted cross-entropy using each set of weights.
  • Figure 5: Confusion matrix using weighted cross-entropy with each set of weights: (a) used conventional cross-entropy, while (b), (c), and (d) employed weighted cross-entropy with the inclusion of weights 6, 7, and 8.