Using event-related brain potentials to help explain posterior-parietal repetition effects in functional magnetic resonance imaging.

Abstract

Background: A major distinction in cognitive psychology is that two different forms of memory exist, explicit and implicit memory. Explicit memory is the effortful, deliberate recollection of past experiences (e.g., remembering a specific event in one’s past). Unlike explicit memory, implicit memory does not require effortful, deliberate recollection of past experiences to influence behavior, yet memory is still expressed (e.g., stimulus processing is facilitated for repeated relative to a non-repeated stimulus, an effect labeled repetition priming,). Each of these forms of memory has been extensively researched and different cortical regions appear to underlie explicit and implicit memory. Recently, however, functional magnetic resonance imaging (fMRI) experiments have indicated that repetition effects in posterior-parietal regions that are usually associated with explicit memory have been measured in implicit memory experiments. Question: Why are repetition effects in posterior-parietal regions observed during an implicit memory task when intentional (or explicit) memory retrieval is not necessary for task performance? Possible Interpretations: Three possible interpretations of these effects were put forward and tested in this experiment using event-related potentials (ERP). Previous studies of posterior-parietal activity suggest that repetition effects in these areas could be attributable to (1) involuntary contamination of explicit memory for previously presented stimuli during putative implicit memory tests, (2) changes in the default mode network (DMN), which is a general network in the brain most active when the participant is not taxed with a cognitive task, and (3) priming in the action pathway of vision. Method: These interpretations were tested using time domain (TD) ERPs collected during a putatively implicit memory experiment that paralleled the procedure used in a preceding fMRI experiment. Time-frequency (TF) analysis also was performed on these ERP data and two frequency bands were extracted (alpha, 8-12 Hz; theta, 2.5-8 Hz). Using TD ERPs and TF analysis allows for arbitrating between the three potential interpretations. Predictions: The explicit memory contamination hypothesis posits that primed (repeated) stimuli should not differ from unprimed (new) stimuli in the TD until 500 ms or later, but afterward there should be a positive deflection (primarily anterior in topography) for primed stimuli compared to unprimed stimuli. Also, the explicit memory contamination posits that anterior alpha or theta power should be greater for primed stimuli than for unprimed stimuli. The change in DMN hypothesis posits that posterior alpha power should be greater for primed stimuli than for unprimed stimuli or anterior theta power should be lesser for primed stimuli than for unprimed stimuli. The priming in the action pathway hypothesis posits that a difference in activity between primed and unprimed conditions should be observed in an early P1 component in occipital temporal locations, and this effect should be greater for highly actionable objects than for nonactionable objects. Results: Primed stimuli elicited a more positive deflection (beginning between 300 and 400 ms) in the TD ERP measures relative to unprimed stimuli, replicating many previous implicit memory ERP studies. One of the three possible interpretations to explain posterior-parietal activation was supported with the ERP and TF analysis. In line with the change in DMN hypothesis, lesser anterior theta power was measured for the primed stimuli relative to the unprimed stimuli. Conclusion: In conclusion, the current ERP experiment offers insight into posteriorparietal repetition effects measured with fMRI during putatively implicit memory experiments. Results supported the hypothesis that the DMN is engaged by primed and unprimed stimuli to different degrees. This will allow researchers to interpret posteriorparietal repetition effects measured with fMRI in a putatively implicit memory experiment as potentially reflecting a change in DMN processing.