Topographic organization of projections from the amygdala to the visual cortex in the macaque monkey.

The topography of amygdaloid projections to the visual cortices in the macaque monkey was examined by injecting the fluorescent tracers Fast Blue and Diamidino Yellow at different locations in the occipital and temporal lobes and mapping the distribution of retrogradely labeled cells in the amygdala. Injections involving regions from rostral area TE to caudal area V1 all resulted in labeled cells within the basal nucleus of the amygdala. Relatively few double-labeled cells were observed even when the two injections were separated by less than 3 mm. The projections were rostrocaudally organized such that projections to caudal visual areas originated from dorsal and caudal portions of the magnocellular division of the basal nucleus while projections to more rostrally situated visual areas originated in more rostral and ventral portions of the basal nucleus. When injections involved rostral and medial portions of area TE, retrogradely labeled cells were observed in the accessory basal and lateral nuclei in addition to the basal nucleus. These data confirm that the amygdala gives rise to feedback projections to all levels of the "ventral stream" visual pathway. The projections do not appear to be diffusely distributed since few double-labeled cells were observed. The largest cells of the basal nucleus, those located in the magnocellular division, project the farthest in the visual system and innervate all occipital and temporal levels. The smaller cells, in the intermediate and parvicellular regions, project to more rostral and medial portions of the visual cortex. These results suggest that the amygdala may have substantial modulatory control over sensory processing at all stages of the ventral-stream visual cortical hierarchy.

Memory lost and regained following bilateral hippocampal damage.

We present a longitudinal neuropsychological study (31 examinations over a period of 18 months) of patient DE DF demonstrated bilateral atrophy of the hippocampal formation and globus pallidus resulting from carbon monoxide poisoning. Eighteen months after the event, the volume of the hippocampal formation was reduced by 42% on the left side and 28% on the right. The patient initially presented with a severe global amnesia. Then, he showed a gradual, yet selective recovery of episodic memory function. Verbal free recall and spatial memory performance remained reduced, whereas immediate word recall and recognition memory, as well as picture learning and memory, improved to levels at the lower range of normal performance. Interestingly, nonspatial associative learning was never much impaired and recovered completely by the end of testing. These data are taken as evidence that the human hippocampal formation does not equally support different forms of episodic memory.