A functional, behavioral, and model-based investigation of human visual memory
Our ability to remember the events of our lives relies upon the formation of associations among the “what”, “where”, and “when” components of the event. Decades of research have focused on elucidating the contributions of subregions of the medial temporal lobe to memory for events. The results of this work converged on the notion that it would be fruitful to investigate the representation of information within subregions of the medial temporal lobe. Specifically, there is consensus that the hippocampus sits at the apex of a cortical circuit, which gives it unparalleled access to all aspects of event processing. In contrast, subregions of the adjacent medial temporal lobe cortex—namely, perirhinal and parahippocampal cortex—are hypothesized to be involved in the representation of object (“what”) and contextual (“where”) aspects of events, respectively. Recent studies have suggested that retrosplenial cortex, a region that has been largely absent from memory models, is also necessary for memory for events and that it is functionally related to parahippocampal cortex.
In four functional magnetic resonance imaging experiments, we investigated the hypothesis that parahippocampal cortex and retrosplenial cortex are preferentially involved in the representation of contextual information and that perirhinal cortex is preferentially involved in the representation of object information. Overall, the results of our experiments support our hypothesis; however, our results are incompatible with a simple functional dissociation between these regions. Furthermore, our results provide evidence for the influence of low-level stimulus features in the representation of contexual information, suggesting that future research should aim to further investigate invariant context representation.
In a behavioral and model-based experiment, we investigated how memory changes as a result of healthy aging. Previous research has suggested that healthy aging is accompanied by an impaired ability to form stimulus-stimulus associations (e.g., “what-where”) with a relative sparing of familiarity for the “what” component of events. An alternative, but not mutually exclusive, hypothesis is that healthy aging is accompanied by an impaired ability to encode stimulus features. Our results support the latter hypothesis, thus providing an alternative framework for the investigation of cognitive aging.