Functional neuroanatomy of amnesia: positron emission tomography studies

In this article, we review the principles of, and provide examples for, the new approach of functional neuropsychology in the field of amnesia. In the permanent amnesic syndrome, positron emission tomography (PET) can provide statistical maps of the brain regions with significantly impaired resting metabolism in comparison with control subjects. These regions include not only Papez's circuit but also the left supramarginal gyrus, which may explain in part the retrograde amnesia present in most cases of amnesic syndrome. This approach is also of great interest in transient global amnesia (TGA) because the defect of episodic memory is highly selective and occurs without permanent damage. The few available PET studies in TGA suggest the dysfunction of a distributed network including the hippocampal region and the prefrontal cortex, with a different pattern individually. Further studies will be necessary to better understand the relationships between the precise cognitive deficits in TGA and the pattern of brain hypometabolism. In Alzheimer's disease (AD), the study of the correlations between memory test scores and metabolic values across a sample of subjects provides a map of those brain structures whose synaptic pathology dysfunction underlies the particular neuropsychological alteration. The distribution of the sites of correlations shows striking differences according to each memory system. This approach should open the way for the unravelling of the neurobiological substrates of both cognitive impairment and compensatory mechanisms in neurodegenerative diseases. Over and above their applications in neurological research, such studies in brain-diseased subjects are particularly useful for establishing cognitive and neurobiological models of human memory, because they allow the highlighting of the neural networks that are essential for memory function. From a cognitive neuroscience perspective, the functional neuropsychology of amnesia is, therefore, complementary to the classic activation paradigm in normal subjects, which identifies the cerebral structures that are involved with, but not necessarily indispensable for, the execution of the task.