In this chapter we review the relatively recent effort on the part of
neuroscientists to use computational
neural network modeling to investigate the
neural basis of subjective
tinnitus. There are advantages and challenges in using a modeling framework to understand this complex auditory disorder. The foremost challenge to modeling a subjective condition such as
tinnitus is the evaluation of the occurrence of
tinnitus in the model. We propose comparing measures of the model's activities (simulated
neuronal activity, behavioral activity, or
neuroimaging activity) with experimental data obtained from studies of
tinnitus in humans and animals; strong agreement with experimental data will provide support for the validity of the simulation of
tinnitus in a particular model. A major advantage of
neural network modeling is that it allows experimentation not possible in animals. Models make it possible to evaluate the contribution of different
neural mechanisms affecting
tinnitus in a principled manner. A model makes predictions that can be tested by experiments thus leading to the designing of focused experiments. We review several
neural models of
tinnitus and discuss published findings from simulations using these models. We conclude with a proposed scheme for investigating
tinnitus that combines
neural network modeling with
brain imaging experiments.
