CONNECTIONIST SIMULATION IN NEUROPSYCHOLOGY

Introduction. The expression 'connectionist neuropsychology' has been applied since the early 1990s to designate an emerging area of research that uses artificial neural networks in an attempt to simulate some of the cognitive disorders that result from different kinds of brain injury, Although some of the models generated by this type of simulation offer a number of advantages over the classic models of information processing, this approach is still not very well known among Spanish researchers and health professionals. Aims. To make this important research tool more widely known and to review the advantages and shortcomings of a series of particular models. Development. After a brief introduction to the fundamental principles of connectionist simulation, some of the most important models involving aphasias, dyslexias, agnosias and apraxias are outlined. Conclusions. Despite their limitations, the models proposed by connectionist neuropsychology seem to be far more accurate and specific them the classic ones. Furthermore, they are easy to handle and make it possible to get much closer to the complex reality of these disorders. This type of research allows different kinds of brain injury to be modelled while also enabling researchers to explore brain functions that have remained unknown tip till now, either for ethical or practical reasons. It there re represents a source of inspiration both for designing experimental research studies and for developing new rehabilitation procedures. [REV NEUROL 2009; 48: 317-21]