The KIII model of the chaotic dynamics of the olfactory system was designed to simulate pattern classification required for odor perception. It was evaluated by simulating the patterns of action potentials and EEG waveforms observed in electrophysiological experiments. It differs from conventional artificial neural networks in relying on a landscape of chaotic attractors for its memory system and on a high-dimensional trajectory in state space for virtually instantaneous access to any low-dimensional attractor. Here we adapted this novel neural network as a diagnostic tool to classify normal and hypoxic EEGs.