Anatomical Model of VOR Using Fuzzy Neural Network

An anatomical model is proposed to model the command of saccadic eye movements. This model of movement control is that its structure is consistent with general anatomy, cell connectivity and functioning of the cerebellar pathways, i.e. motor orders processing, to compute an approximate inverse function of the bio-mechanical function of eye and head sensors. Its parameters are deduced from a stability analysis constraint and system identification to produce desired output. In this research, a predicting element, necessary for inversion and thus for movement accuracy, is modeled by an artificial neural network which is an Adaptive Neuro Fuzzy Inference System (ANFIS) with online learning. According to anatomy, in this model, in the first step, the proposed ANN is trained with a slow movement in a closed-loop pathway to initiate a primary structure. In the next steps, the model is able to accurately simulate saccadic eye movements and keeps the online learning to reduce the error. Experiment results show that the proposed model has been stable in different saccadic movements and error tends to neglected values.