CCN Colloquium: "Episodic Memory Networks and Representations"
Roberto Cabeza, PhD (Duke University)
This talk presents fMRI evidence from our lab regarding episodic memory networks and representations in young and older adults. Networks are investigated using functional connectivity and graph theory, and representations, with representational similarity analyses. The network data indicate that episodic memory success is associated with stronger hippocampal interactions with the rest of the brain, and particularly with a "hippocampal assembly" of occipito-temporal, ventral parietal, and frontal regions. In older adults, medial temporal lobe interactions with the rest of the brain are attenuated while prefrontal cortex interactions with the rest of the brain are enhanced, in a form of compensation. The representational data show that encoded representations predict subsequent memory depending not only on their perceptual vs. conceptual nature, but also on where in the brain they are stored, possibly reflecting an interaction between representations and networks. In older adults, sensory representations in more posterior temporal regions are dedifferentiated, whereas semantic representations in more anterior temporal regions are hyperdifferentiated, with this last effect associated with compensation. The talk ends with two new studies linking functional connectivity and representations.
Medicine, Natural Sciences, Lecture/Talk, Panel/Seminar/Colloquium, Research