A neural network as an approach to clinical diagnosis.

During the past decade, artificial neural networks have been established as promising psychological and computational models. The proponents of neural computing believe that it offers new solutions to problems that have been intractable so far. To study the suitability of neural networks for performing sequential diagnostic classification, I have used a network that, over time, becomes increasingly proficient at diagnosing dementia. A description of the implementation, training, and behavior of this network illustrates how neural-network technology might contribute to clinical computing.