ACTIVE DISTURBANCE REJECTION CONTROL AND NEURAL MASS MODEL FOR EEG SEIZURE PROGNOSIS
Abhinav Kar, Rishab Das, Aayush Gupta, A. Sharmila
School of Electrical Engineering, Vellore Institute of Technology,VIT University, Vellore, Tamil Nadu – 632014, India
ABSTRACT: In this research, EEG signals were used to simulate and monitor the brain’s electrical activity, especially focusing on detecting epileptic seizures. The study uses the Active Disturbance Rejection Control (ADRC) method as a control strategy, known for its ability to control complex and uncertain systems. This approach excels at estimating and mitigating perturbations in the brain’s electrical activity and offers a robust solution for seizure control. A comparative analysis with the traditional proportional integration (PI) method revealed that the ADRC exhibits superior performance. This conclusion is underlined by the significantly lower values of key error metrics such as mean absolute error (MAE) and root mean square error (RMSE) obtained with ADRC. These findings confirm the enhanced efficacy and accuracy of ADRC in stabilizing brain activity during epileptic episodes and mark it as a promising avenue for advanced neurological applications.
INDEX TERMS: EEG signals, Active Disturbance Rejection Control (ADRC), Neural Mass Model (NMM), Comparison with PI control