Neural network modeling of central auditory dysfunction in Alzheimer's disease

Abstract

Background: Many reports have described that individuals with Alzheimer's disease show neural breakdown in the brainstem nuclei, hippocampus and auditory cortex in the temporal lobe of the brain. However, it is still unclear whether auditory skills mediated by these areas differ across individuals with and without Alzheimer's disease and how these auditory skills are further confounded by reduction in cognitive function in individuals with AD.

Aim: The aim of this study is to discover the hidden and nonlinear associations among higher-order auditory and cognitive processes in individuals with and without Alzheimer's disease through artificial neural network analysis.

Methods: The analyses were based on auditory test data obtained from nine clinically confirmed Alzheimer's disease cases and nine age-matched controls. Hearing threshold sensitivity was equivalent across groups, indicating similar peripheral auditory function. Auditory function was evaluated by standardized tests of auditory closure, auditory attention, and auditory figure ground in listeners with and without Alzheimer's disease. The inputs used for analyses were cognitive status and auditory function. The dependent variables were RAU scores computed from scores of auditory tests.

Results: Artificial neural networks showed a complex relationship between the input variables (cognitive status and auditory function) that cannot be predicted simply on the basis of cognition differences between individuals with and without Alzheimer's disease.

Conclusion: The results of this study suggest that central auditory function declines with age, regardless of changes in cognitive function.