ABSTRACT

The amnesia syndrome is characterized by normal perceptual, linguistic and intellectual functioning together with an inability to remember explicitly recent events and new information. Amnesic patients generally exhibit spared priming effects on implicit memory tasks, despite poor explicit memory. An amnesia model and an artificial neural network simulation framework for the study of the amnesia syndrome is proposed. The network is combined by modules responsible for the input receiving, short-term and long-term memory. Each module of the network is combined by artificial neurons that represent neural assemblies exhibiting firing pattern sequences corresponding to memory objects. The parameters of the network are set so, that the performance of the network is similar to that of normal humans. We damaged the information preserving capacity of the short-term memory module, measured the effects and compared the performances to those of the humans with amnesia syndrome. An amnesia model and an artificial neural network simulation framework for the study of the amnesia syndrome is proposed. We damaged the information preserving capacity of the short-term memory module, measured the effects and compared the performances to those of the humans with amnesia syndrome.