Classification of EEG Signals Based on Sparrow Search Algorithm-Deep Belief Network for Brain-Computer Interface

Area of Science:

• Neuroscience
• Computer Science
• Biomedical Engineering

Background:

• Brain-computer interfaces (BCI) face challenges in accurately recognizing motor imagery (MI) brain signals.
• Existing classification methods for MI require improvement in accuracy and robustness.

Purpose of the Study:

• To develop an advanced classification method for MI-BCI systems.
• To enhance the accuracy and robustness of MI signal recognition using an optimized deep learning model.

Main Methods:

• Empirical Mode Decomposition (EMD) was used to extract EEG features.
• A Deep Belief Network (DBN) was optimized using the Sparrow Search Algorithm (SSA), creating the SSA-DBN model.
• The SSA-DBN model's performance was evaluated on two public and one private dataset.

Main Results:

• The SSA-DBN method demonstrated superior classification accuracy compared to baseline methods.
• On a private dataset, SSA-DBN achieved 87.83% accuracy, a 10.38% improvement over standard DBN.
• Significant accuracy improvements were also observed on the BCI IV 2a (86.14%) and SMR-BCI (87.21%) datasets.

Conclusions:

• The SSA-DBN model offers enhanced classification capabilities for MI-BCI.
• This approach shows potential for advancing the field of brain-computer interfaces through improved signal recognition.