Comprehensive analysis of prefrontal cortex-directional rhythms categorization for rehabilitation

Area of Science:

• Neuroscience
• Biomedical Engineering
• Machine Learning

Background:

• Prefrontal Cortex-Directional Rhythms (PFC-DR) classification is vital for Brain-Computer Interface (BCI) systems.
• Effective BCI systems are crucial for rehabilitating individuals with voluntary movement impairments.

Purpose of the Study:

• To examine traditional classification techniques in BCI.
• To highlight the efficacy of Radial Basis Function-Support Vector Machine (RBF-SVM) approaches for PFC-DR classification.
• To introduce a generalized RBF-SVM classifier for enhanced BCI performance.

Main Methods:

• Utilized a generalized Radial Basis Function-Support Vector Machine (RBF-SVM) classifier.
• Employed a 10-fold repeated random train-test split cross-validation technique.
• Performed confusion matrix analysis to validate classification accuracy.

Main Results:

• The generalized RBF-SVM classifier achieved a high overall accuracy of 96.91%.
• This performance surpasses existing machine learning-based approaches for PFC-DR classification.
• Confusion matrix analysis confirmed the classifier's effectiveness.

Conclusions:

• The generalized RBF-SVM approach offers a robust and accurate method for PFC-DR classification in BCI systems.
• This advancement holds significant potential for improving rehabilitation outcomes in patients with movement disorders.
• The study validates the superiority of the proposed RBF-SVM method over traditional techniques.