A NEURAL MODEL OF MEMORY IMPAIRMENT IN DIFFUSE CEREBRAL ATROPHY

Background. Computer-supported neural network models have been subject ed to diffuse. progressive deletion of synapses/neurons, to show that modelling cerebral neuropathological changes can predict the pattern o f memory degradation in diffuse degenerative processes such as Alzheim er's disease. However, it has been suggested that neural models cannot account for more detailed aspects of memory impairment, such as the r elative sparing of remote versus recent memories. Method. The latter c laim is examined from a computational perspective, using a neural asso ciative memory model. Results. The neural network model not only demon strates progressive memory deterioration as diffuse network damage occ urs, but also exhibits differential sparing of remote versus recent me mories. Conclusions. Our results show that neural models can account f or a large variety of experimental phenomena characterising memory deg radation in Alzheimer's patients. Specific testable predictions are ge nerated concerning the relation between the neuroanatomical findings a nd the clinical manifestations of Alzheimer's disease.