چکیده:
Handwriting is a complex cognitive–motor process involving the integration of sensorimotor coordination, visual processing, motor planning, and higher cognitive functions. Electroencephalography (EEG) has emerged as a valuable noninvasive neuroimaging technique for investigating the neural mechanisms underlying handwriting and motor imagery tasks due to its high temporal resolution and portability. Recent advances in brain–computer interface (BCI) systems, signal processing techniques, and deep learning algorithms have significantly improved the decoding and interpretation of EEG signals associated with handwriting-related activities.
This review summarizes recent developments in EEG-based handwriting and motor imagery research, with a particular focus on neural oscillations, signal preprocessing, feature extraction, classification approaches, and deep learning architectures. Major EEG rhythms associated with handwriting, especially mu and beta bands, are discussed in relation to sensorimotor brain networks. Traditional machine learning approaches such as Common Spatial Pattern (CSP), Linear Discriminant Analysis (LDA), and Support Vector Machines (SVM) are compared with modern deep learning methods including convolutional neural networks (CNNs), recurrent neural networks (RNNs), EEGNet, and hybrid CNN–RNN models.
Furthermore, the review highlights current challenges including signal noise, inter-subject variability, limited dataset availability, and poor cross-session generalization. Emerging research directions such as interpretable artificial intelligence, multimodal neuroimaging integration, transfer learning, and portable real-time EEG systems are also discussed. The findings demonstrate that EEG-based handwriting analysis represents a promising interdisciplinary field with important applications in neurorehabilitation, cognitive assessment, assistive technologies, and intelligent BCI systems.
This review summarizes recent developments in EEG-based handwriting and motor imagery research, with a particular focus on neural oscillations, signal preprocessing, feature extraction, classification approaches, and deep learning architectures. Major EEG rhythms associated with handwriting, especially mu and beta bands, are discussed in relation to sensorimotor brain networks. Traditional machine learning approaches such as Common Spatial Pattern (CSP), Linear Discriminant Analysis (LDA), and Support Vector Machines (SVM) are compared with modern deep learning methods including convolutional neural networks (CNNs), recurrent neural networks (RNNs), EEGNet, and hybrid CNN–RNN models.
Furthermore, the review highlights current challenges including signal noise, inter-subject variability, limited dataset availability, and poor cross-session generalization. Emerging research directions such as interpretable artificial intelligence, multimodal neuroimaging integration, transfer learning, and portable real-time EEG systems are also discussed. The findings demonstrate that EEG-based handwriting analysis represents a promising interdisciplinary field with important applications in neurorehabilitation, cognitive assessment, assistive technologies, and intelligent BCI systems.
| بازنشر اطلاعات | |
 |
این مقاله تحت شرایط Creative Commons Attribution-NonCommercial 4.0 International License قابل بازنشر است. |
