Mean Membrane Potential Estimation for Neural Mass Models in EEG Recordings Using a Linear State Observer

The paper presents a classical Kalman filter design to estimate the hidden states of a cortical column. The method is based on a linearized mathematical description of a macrocolumn and its EEG measurements. The nonlinear neural mass model is built using a convolutional-based model which describes the intrinsic connection between its neuronal populations. Simulation results are provided to indicate the reliability of the estimates delivered by the proposed method in scenarios considering the trade-off between the model accuracy and measurement noise.