Improved Assessment of Vestibular Systems by Fuzzy and Neural Hybrid Modeling Techniques

A fuzzy logic and artificial neural network (ANN) hybrid software model of vestibular system function will be investigated to improve assessment of the vestibular systems. Recent advances in fuzzy logic and ANNs have provided new modeling techniques to narrow the gap between the model and real physiology systems. Specifically, hierarchical fuzzy systems and other fuzzy logic modeling techniques will be investigated and applied to improve eye muscle and vestibular modeling accuracy. Temporal and/or asynchronous ANNS will be incorporated for enhancing the vestibular-ocular reflex (VOR) neuronal pathway model. Utilizing hierarchical fuzzy logic systems adds benefit to tranditional software moedling because as future physiological properties are discovered the model can be adapted to include these aspects. Further, by utilizing temporal and/or asynchronous properties, the ANN based neuronal pathway model will be closer to real physiological systems. Therefore, the Phase I effort will determine the feasible improvement in the modeling of the vestibular system, eye muscle, and neuronal pathway. In addition, a final model which encompasses the above three will be outlined for prototypes in the Phase II. The purpose is to enable a clinician/scientist to match VOR signal abnormalities to physiological changes without the need for prior "examples".