Neural activity in the hippocampus predicts individual visual short-term memory capacity
Article first published online: 20 APR 2013
Copyright © 2013 Wiley Periodicals, Inc.
Volume 23, Issue 7, pages 606–615, July 2013
How to Cite
von Allmen, D. Y., Wurmitzer, K., Martin, E. and Klaver, P. (2013), Neural activity in the hippocampus predicts individual visual short-term memory capacity. Hippocampus, 23: 606–615. doi: 10.1002/hipo.22121
- Issue published online: 21 JUN 2013
- Article first published online: 20 APR 2013
- Accepted manuscript online: 20 MAR 2013 06:27AM EST
- Manuscript Accepted: 6 MAR 2013
- Swiss National Science Foundation. Grant Number: CR13I1_127115
- posterior parietal cortex;
- visual short-term memory;
- memory capacity;
- functional magnetic resonance imaging
Although the hippocampus had been traditionally thought to be exclusively involved in long-term memory, recent studies raised controversial explanations why hippocampal activity emerged during short-term memory tasks. For example, it has been argued that long-term memory processes might contribute to performance within a short-term memory paradigm when memory capacity has been exceeded. It is still unclear, though, whether neural activity in the hippocampus predicts visual short-term memory (VSTM) performance. To investigate this question, we measured BOLD activity in 21 healthy adults (age range 19–27 yr, nine males) while they performed a match-to-sample task requiring processing of object-location associations (delay period = 900 ms; set size conditions 1, 2, 4, and 6). Based on individual memory capacity (estimated by Cowan's K-formula), two performance groups were formed (high and low performers). Within whole brain analyses, we found a robust main effect of “set size” in the posterior parietal cortex (PPC). In line with a “set size × group” interaction in the hippocampus, a subsequent Finite Impulse Response (FIR) analysis revealed divergent hippocampal activation patterns between performance groups: Low performers (mean capacity = 3.63) elicited increased neural activity at set size two, followed by a drop in activity at set sizes four and six, whereas high performers (mean capacity = 5.19) showed an incremental activity increase with larger set size (maximal activation at set size six). Our data demonstrated that performance-related neural activity in the hippocampus emerged below capacity limit. In conclusion, we suggest that hippocampal activity reflected successful processing of object-location associations in VSTM. Neural activity in the PPC might have been involved in attentional updating. © 2013 Wiley Periodicals, Inc.