The assumptions underlying neuroimaging, and problems in its analysis and i
nterpretation, are commonly underestimated in neuropsychology. The ways in
which fluoro-deoxy-glucose (FDG) positron emission tomography (PET) data ca
n be analysed are discussed. PET findings from four patients who had suffer
ed severe amnesia, following episodes of acute hypoxia, are presented. Thes
e patients had shown evidence of medial temporal (hippocampal and parahippo
campal) atrophy on MRI brain scans. The PET data were analysed in several d
ifferent ways. The converging findings were that the patients showed bilate
ral thalamic hypometabolism, and there was also evidence of retrosplenial h
ypometabolism bilaterally, Cognitively, these patients performed most like
other patients with medial temporal lesions, but the results indicate that
structural lesions can have distal metabolic effects on structures elsewher
e. These findings are interpreted in the light of neuroanatomical observati
ons concerning parallel projections between medial temporal lobe structures
and the thalamus, some of which pass via the retrosplenium.