EOG-based eye movement recognition using GWO-NN optimization.

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction
• Artificial Intelligence

Background:

• Human-computer interface (HCI) systems are increasingly controlled by electrooculography (EOG) signals.
• Recognizing eye movement patterns from EOG signals allows for control of HCI systems and task execution.

Purpose of the Study:

• To propose a novel model for eye movement recognition using EOG signals.
• To enhance HCI control through accurate EOG signal pattern recognition.

Main Methods:

• Utilized Grey Wolf Optimization (GWO) to minimize the error function of a neural network (NN) classifier.
• Developed a new model for eye movement recognition based on EOG signals.
• Compared the proposed model's performance against conventional methods.

Main Results:

• The proposed GWO-NN model demonstrated superior performance compared to conventional methods.
• Error minimization analysis confirmed the effectiveness of the GWO algorithm.
• Recognition performance analysis showed high accuracy, sensitivity, specificity, precision, and F1 score, with low FPR, FNR, NPV, and FDR.

Conclusions:

• The novel GWO-NN model offers a significant advancement in EOG-based eye movement recognition for HCI control.
• The method achieves robust performance across various metrics, indicating its reliability and effectiveness.
• This approach provides a promising direction for developing more intuitive and efficient HCI systems.