Objective: Although some of the anatomical underpinnings of learning and memory systems have been identified, there remains little understanding of how the brain moves from acquiring new information to retaining it. This study was designed to further explore and elucidate the neural mechanisms underlying encoding and memory in a common real-life task, that is, face-name associations. One possible outcome is that the tasks will recruit different neural structures mediating these processes, which can be identified through contrast analysis of activations. Alternatively, it is possible that similar anatomical regions, such as the hippocampus and parahippocampal gyrus, may be involved in both tasks. In that case, analysis of blood oxygenation level dependent (BOLD) amplitude differences between the tasks in those common neural structures may be able to detect whether physiological activation differences occur in encoding versus memory.
Methods: Five healthy adult participants underwent high-field magnetic resonance imaging (MRI) while learning face-name pairs (encoding phase) and during a multiple-choice recognition task after a brief delay (memory phase). Average activation and BOLD response amplitudes in specific regions of interest and whole-brain activation maps were analyzed.
Results: Common activations were observed in the encoding and recognition memory tasks in several regions of interest encompassing the medial temporal and inferior occipital regions. However, higher BOLD response amplitudes occurred in the right fusiform gyrus and the right hippocampus during encoding. In contrast, higher amplitudes were detected in the lingual gyrus bilaterally during recognition memory. Encoding activated distributed prefrontal and temporal cortical regions bilaterally, which mediate attentional, executive, language, and memory systems. Recognition memory recruited a different network of regions encompassing convergence zones in the left prefrontal cortex and the parietal-occipital-temporal region bilaterally, where multimodal visual association, language, memory, and decision-making systems interact.
Conclusions: Higher BOLD response amplitudes in the right fusiform gyrus and the right hippocampus during face-name encoding suggest a potentially specific binding pathway where disparate information might be neurally linked. In contrast, the increased BOLD response in the lingual gyrus during recognition memory may indicate a key neural substrate for memory consolidation and long-term knowledge of what is learned. Whole-brain activation maps revealed task-specific differences in areas of the prefrontal, temporal, and occipital-parietal-temporal junctions as well. Findings suggest that there are distinctive anatomical and physiological nodes for face-name learning and memory within large-scale cortical-subcortical networks. Hence, lesions in fairly widespread cerebral regions may potentially disrupt specific binding and/or memory consolidation processes.