Synchronization Control Of Ephaptic Coupled Neurons Based On Unscented Kalman Filter (UKF)

Epilepsy is a kind of chronic diseases of the nervous system with the dysfunctionof central nervous system caused by neuronal paradoxical firing. Drugs and surgeryare the main treatment for epilepsy. However,30%of patients with epilepsysymptoms didn’t relieve under these therapies. For many years epilepsy has been oneof the focuses of clinical problem. Recently electromagnetic stimulation makesencouraging progress in controlling epilepsy and seizures. The synchronization amongneurons and neuronal population is an important mechanism that brains encode andprocess information, abnormal synchrony (excessive-synchrony or hyper-synchrony)of neuronal network may lead to mental disorders such as epilepsy, so it is of greatpractical significance to study the desynchronizing mechanism of neuronal network.Since nervous system is a highly complicated nonlinear system, the establishment ofthe neuron model provides the possibility for us to study brain activity, but in actualphysical experiment the parameters in these neuron models are too hard to measure.The emergence of Unscented Kalman Filter (UKF) provides an effective way toestimate the neuron’s state which contains a lot of noise, it can estimate the state andparameters of nonlinear system, make it possible to measure and control the neuronmodel.In this paper we firstly establish a ephaptic coupled neuronal network modelembedded in four neurons based on hippocampal pyramidal neuron model, analyzethe dynamic properties of this model, and then explore the parameters which influencethe synchronous feature of neuronal network model and find the key parameters inthis model, then estimate the parameters and state variables in the model from themeasurement of action potential containing a lot of noise with UKF.The next part is the estimation and control of the key parameters in the model.First of all we estimate the unmeasured key parameters in the ephaptic coupledneuronal network model with UKF, then constitute a feedback control systemcombining UKF and closed-loop control, control the key parameter of the modelthrough the adjustment of stimulus strength, thus change the synchronous state of theneuronal network model, at last realize desynchronization of the neuronal networkmodel. In this paper we combine closed-loop control strategy with UKF, put forward anew closed loop control strategy which can solve the control problem whenparameters are unknown and guarantee robustness of the system. This paper presentsthe detailed design process and steps, the simulation results show the effectiveness ofthe designed control algorithm.The results of this study provide theoretical basis for epileptic electricalstimulation mechanism.