Review
Prior perceptual processing enhances the effect of emotional arousal on the neural correlates of memory retrieval

https://doi.org/10.1016/j.nlm.2013.12.012Get rights and content

Abstract

A fundamental idea in memory research is that items are more likely to be remembered if encoded with a semantic, rather than perceptual, processing strategy. Interestingly, this effect has been shown to reverse for emotionally arousing materials, such that perceptual processing enhances memory for emotional information or events. The current fMRI study investigated the neural mechanisms of this effect by testing how neural activations during emotional memory retrieval are influenced by the prior encoding strategy. Participants incidentally encoded emotional and neutral pictures under instructions to attend to either semantic or perceptual properties of each picture. Recognition memory was tested 2 days later. fMRI analyses yielded three main findings. First, right amygdalar activity associated with emotional memory strength was enhanced by prior perceptual processing. Second, prior perceptual processing of emotional pictures produced a stronger effect on recollection- than familiarity-related activations in the right amygdala and left hippocampus. Finally, prior perceptual processing enhanced amygdalar connectivity with regions strongly associated with retrieval success, including hippocampal/parahippocampal regions, visual cortex, and ventral parietal cortex. Taken together, the results specify how encoding orientations yield alterations in brain systems that retrieve emotional memories.

Introduction

It is well known that emotionally arousing events are better remembered than neutral events. This effect is modulated by the release of stress hormones that affect noradrenergic transmission in the basolateral amygdala and its interactions with the anterior medial temporal lobes, which are important for memory formation, consolidation, and/or retrieval operations (Dolcos et al., 2004, Kensinger, 2009, LaBar and Cabeza, 2006, McGaugh, 2004, Wolf, 2008). However, very little is known about the neural bases of cognitive factors known to modulate emotional memory, such as encoding orientations that determine how emotional information is processed.

One variable that may modulate the effect of emotional arousal on memory is the degree to which perceptual properties of stimuli or events are attended or processed. A fundamental idea in memory research is that semantic processing leads to better memory than perceptual processing (the levels-of-processing effect, or LOP) (Craik & Lockhart, 1972). However, the vast majority of the studies showing the LOP effect have used affectively neutral information (Craik, 2002), and it is unclear if the effect generalizes to emotional information. In fact, studies have shown that attending to and processing perceptual, rather than semantic, aspects of incoming information tends to boost emotional memory (Jay et al., 2008, Reber et al., 1994, Ritchey et al., 2011). For instance, Reber et al. (1994) found that whereas recall of neutral words showed the standard LOP effect, such that performance was higher for semantic than perceptual processing, recall of emotional words showed a “reverse LOP effect,” wherein performance was higher for perceptual than semantic processing. More generally, mounting evidence from different literatures and methodological techniques suggests a strong link between emotional memory and perceptual processing, particularly for negative emotions. For example, vivid visual imagery of negative events is frequently reported by patients with post-traumatic stress disorder (Hackmann & Holmes, 2004). In addition, emotional memories are often subjectively rated as more vivid (e.g., Kensinger and Corkin, 2003, Ochsner, 2000) and may contain sensory details absent in neutral memories (Doerksen & Shimamura, 2001). In the domain of implicit memory, emotion has been shown to enhance perceptual (LaBar et al., 2005) but not conceptual (Ramponi, Handelsman, & Barnard, 2010) priming effects. Finally, research on affect labeling (e.g., Lieberman et al., 2007) has shown that activity in emotion-related limbic regions is reduced when people attend to or label the semantic category or emotional pictures, suggesting that semantic encoding orientations may actually dampen emotional experience.

The neural mechanisms whereby perceptual processing enhances emotional memory are unclear. There is evidence that emotion enhances visual cortex activity (Bradley et al., 2003, Lang et al., 1998, Sabatinelli et al., 2007, Simpson et al., 2000), most likely via feedback projections from the amygdala to visual cortex (Amaral et al., 2003, Anderson and Phelps, 2001). However, neural evidence of the link between perceptual processing and emotional memory is scarce and available only for encoding. One recent fMRI study (Todd, Talmi, Schmitz, Susskind, & Anderson, 2012) found that visual noise overlaid on scenes was rated as less noisy when the scenes were emotional, an effect that was linked to activations in the amygdala and occipital cortex. Importantly, these activations also predicted memory vividness ratings on a recognition test 1 week later (see also Todd, Schmitz, Susskind, & Anderson, 2013). In another study (Kensinger, Garoff-Eaton, & Schacter, 2007), right amygdalar activity during object encoding predicted the ability to distinguish between identical and visually similar objects during retrieval. Finally, Ritchey et al. (2011) found that right amygdala activity predicted subsequent emotional memory to a greater extent when pictures were perceptually- rather than semantically-encoded.

Although the foregoing studies established preliminary links between perceptual processing and emotional memory encoding, it is an open question whether prior perceptual processing enhances the neural mechanisms of emotional memory during retrieval. Because perceptual processing occurs during encoding, any effect on retrieval mechanisms would have to be mediated by the nature of memory representations. That is, a difference in amygdalar activity as a function of prior processing strategy cannot be explained by processes elicited by the retrieval cue, but rather has to be mediated by the nature of the memory representations stored during perceptual processing and recovered during retrieval. It is currently uncertain how perceptual processing applied during memory formation will affect neural activations that occur during later emotional retrieval.

A second open question is whether prior perceptual processing enhances the quality of emotional memories. In addition to quantitative differences in memory strength, episodic memory researchers distinguish two qualitatively different forms of memory: recollection and familiarity (for review, see Yonelinas, 2002). Recollection refers to remembering a past event together with its associated contextual details, whereas familiarity refers to knowing that the event occurred in the past in the absence of contextual details. Several studies have shown that emotion tends to enhance recollection rather than familiarity (Dougal et al., 2007, Ochsner, 2000, Sharot et al., 2007). Dolcos, LaBar, and Cabeza (2005) found that, during retrieval, the emotion-induced enhancement of recollection was mediated by increased activity in the amygdala and the hippocampus. It is unknown, however, how the mechanisms of emotional recollection are modulated by prior perceptual processing.

Finally, a third open question is whether the enhancing effects of prior perceptual processing on emotional memory are mediated by localized changes in the amygdala or whether they also involve changes in the interactions between the amygdala and the regions mediating successful memory operations. In fMRI studies of recognition memory, successful retrieval has been associated with increased activations in hippocampus (e.g., Buckner and Wheeler, 2001, Dobbins et al., 2003), and the enhancing effects of emotion on memory have been linked to increases in amygdalar activity with the hippocampus (Dolcos et al., 2005). In non-memory studies, as noted above, emotional effects on visual processing have been linked with feedback projections from the amygdala to visual cortex (Amaral et al., 2003, Anderson and Phelps, 2001). However, it is currently uncertain how prior perceptual processing enhances amygdalar interactions with medial temporal and other cortical regions associated with memory success. One possibility is that perceptual processing during encoding may draw attention to the salient features of the emotional materials, resulting in a richer, more detailed memory trace with more arousal attached to it. As such, when participants see that picture again, the retrieved memory may be more likely to have retained that vividness and arousal and re-engage the amygdala, MTL and visual cortices.

To investigate these three questions, we used the paradigm illustrated by Fig. 1. Emotional and neutral scenes were incidentally encoded under instructions to attend to either semantic or perceptual features during scene viewing. Following each scene, participants were asked to make an arousal rating and to answer a question specific to perceptual or the semantic condition. Two days later, memory for the scenes was tested with a task in which participants rated memory strength and indicated the occurrence of recollection (retrieval of contextual details).

The fMRI analyses focused on the three questions mentioned above. First, to investigate whether prior perceptual processing modulates the neural mechanisms of emotional memory strength, we used participants’ rating of memory strength as a parametric regressor and identified regions where activity increased as a function of emotional memory success. On the basis of the aforementioned evidence linking the amygdala to emotion-induced memory retrieval enhancement (Buchanan, 2007) and effects of perceptual processing on emotional encoding (Ritchey et al., 2011, Todd et al., 2012) we predicted the amygdalar activity associated with emotional memory strength would be enhanced by prior perceptual processing. Second, to investigate whether prior perceptual processing enhances the neural mechanisms of emotional recollection, we compared activity for trials eliciting recollection to trials eliciting strong memories without recollection. On the basis of evidence linking emotional recollection to the amygdala and the hippocampus (Dolcos et al., 2005), we predicted that amygdalar and hippocampal activity associated with emotional recollection would be enhanced by prior perceptual processing. Finally, to investigate whether prior perceptual processing enhances amygdalar interactions with medial temporal and other cortical regions, we used the amygdala as a seed in a functional connectivity analyses. We predicted that prior perceptual processing would enhance amygdalar connectivity with regions strongly associated with retrieval success, such as hippocampal/parahippocampal regions, visual cortex, and ventral parietal cortex.

Section snippets

Participants

Twenty-one young adults participated in the study. All participants were right-handed, native English speakers with no history of psychiatric or neurological illness. Participants provided written informed consent in accordance with the Institutional Review Board of Duke University Medical Center. One participant was excluded for excessive head motion and one was excluded for problems with image acquisition, leaving data from 19 participants included in analysis (9 female; ages 18–29, m = 23.0, SD

Behavioral results

All behavioral results were reported in a previous paper (Ritchey et al., 2011); for convenience only, we reiterate them here. Average arousal ratings at encoding were entered into a repeated-measures ANOVA with factors of emotion (negative, neutral, positive) and task (semantic, perceptual). There was a significant main effect of emotion, F(2, 38) = 206.31, p < .001, ηp2=.92. Follow-up tests revealed that negative pictures were rated as more arousing than neutral, F(1, 19) = 335.38, p < .001, or

Discussion

We investigated neural activations during retrieval of emotional and neutral pictures, which were incidentally encoded 2 days prior under instructions to focus on either perceptual or semantic properties. The study yielded three main findings. First, relative to prior semantic processing, prior perceptual processing enhanced neural activity related to memory strength in the right amygdala for emotional but not neutral pictures. This finding was replicated when pictures were not classified as

Acknowledgments

This work was supported by the National Institutes of Health Grant Nos. NS41328, AG23770, and AG19731. M.R. was supported by National Research Service Award Grant no. F31MH085384 and I.D. was supported by National Research Service Award no. F32AG038298.

References (58)

  • J. Rissman et al.

    Measuring functional connectivity during distinct stages of a cognitive task

    Neuroimage

    (2004)
  • H. Sheridan et al.

    Levels of processing influences both recollection and familiarity: Evidence from a modified remember-know paradigm

    Consciousness and Cognition

    (2012)
  • S.D. Slotnick et al.

    Distinct prefrontal cortex activity associated with item memory and source memory for visual shapes

    Brain Res. Cogn. Brain Res.

    (2003)
  • K.L. Vilberg et al.

    Memory retrieval and the parietal cortex: A review of evidence from a dual-process perspective

    Neuropsychologia

    (2008)
  • O.T. Wolf

    The influence of stress hormones on emotional memory: Relevance for psychopathology

    Acta Psychologica (Amst)

    (2008)
  • A.P. Yonelinas

    The nature of recollection and familiarity: A review of 30 years of research

    Journal of Memory and Language

    (2002)
  • S. Anders et al.

    Brain activity underlying emotional valence and arousal: A response-related fMRI study

    Human Brain Mapping

    (2004)
  • A.K. Anderson et al.

    Lesions of the human amygdala impair enhanced perception of emotionally salient events

    Nature

    (2001)
  • M.M. Bradley et al.

    Activation of the visual cortex in motivated attention

    Behavioral Neuroscience

    (2003)
  • T.W. Buchanan

    Retrieval of emotional memories

    Psychological Bulletin

    (2007)
  • R.L. Buckner et al.

    The cognitive neuroscience of remembering

    Nature Reviews Neuroscience

    (2001)
  • R. Cabeza et al.

    The parietal cortex and episodic memory: An attentional account

    Nature Reviews Neuroscience

    (2008)
  • F.I. Craik

    On the transfer of information from temporary to permanent memory

    Philosophical Transactions of the Royal Society of London Series A-Mathematical Physical and Engineering Sciences

    (1983)
  • F.I. Craik

    Levels of processing: Past, present. and future?

    Memory

    (2002)
  • S. Doerksen et al.

    Source memory enhancement for emotional words

    Emotion

    (2001)
  • F. Dolcos et al.

    Remembering one year later: Role of the amygdala and the medial temporal lobe memory system in retrieving emotional memories

    Proceedings of the National Academy of Sciences of the United States of America

    (2005)
  • S. Dougal et al.

    The role of medial temporal lobe in item recognition and source recollection of emotional stimuli

    Cognitive, Affective, & Behavioral Neuroscience

    (2007)
  • H. Garavan et al.

    Amygdala response to both positively and negatively valenced stimuli

    NeuroReport

    (2001)
  • J.M. Gardiner

    Functional aspects of recollective experience

    Memory & Cognition

    (1988)
  • Cited by (15)

    • Effect of the congruity of emotional contexts at encoding on source memory: Evidence from ERPs

      2022, International Journal of Psychophysiology
      Citation Excerpt :

      Their results revealed that negative rather than positive or neutral pictures interfered with semantic processing tasks (commonness judgments about objects), resulting in slower reaction times, but not perceptual tasks (size judgments about objects). The finding is consistent with the idea that negative noise information may damage semantic processing rather than perceptual processing (Dew et al., 2014; Kensinger, 2009; Sakaki et al., 2011). These results suggest that the differential effect of emotional contexts on source memory may be due to various source details, retrieval of semantic source details, which need semantic processing, were destroyed by emotionally contextual information (Maratos and Rugg, 2001; Ventura-Bort et al., 2016; Ventura-Bort et al., 2020a), and yet retrieval of perceptual source details, such as spatial location or color, were not affected (Jeon et al., 2020), even was enhanced by the emotionally contextual information (Macri et al., 2020; Xie and Zhang, 2017).

    • Suppress to feel and remember less: Neural correlates of explicit and implicit emotional suppression on perception and memory

      2020, Neuropsychologia
      Citation Excerpt :

      The MTL ROI mask was created based primarily on the structural images for bilateral AMY, HC, and PHC from the Automated Anatomical Labeling Atlas (AAL, Tzourio-Mazoyer et al., 2002) in SPM, given our a priori hypotheses regarding the involvement of these regions in emotional memory encoding (Dolcos et al., 2017; Murty et al., 2011). A combination of whole-brain and ROI-based analyses has similarly been employed in recent studies of emotional memory from our group and from others (Dew et al., 2014; Dolcos et al., 2013; Kaneda et al., 2017; Shafer and Dolcos, 2012, 2014). Also following previous studies from our group (Dolcos et al., 2013; Shafer and Dolcos, 2014), these ROI images were used in conjunction with an in-house AMY mask3 to correct for inaccurate spatial coverage provided by the AAL mask.

    • The way to her heart? Response to romantic cues is dependent on hunger state and dieting history: An fMRI pilot study

      2015, Appetite
      Citation Excerpt :

      Both groups were significantly more responsive to romantic stimuli after a meal than prior to it, with more limited differences in activity in HDs and NDs. NDs showed greater activation when fed than fasted in response to romantic cues in the hippocampus and amygdala, areas associated with emotional memory (Dew, Ritchey, LaBar, & Cabeza, 2014); the putamen, linked to reward (Elliott, Newman, Longe, & Deakin, 2003); the middle temporal gyrus, correlated with facial recognition (Leveroni et al., 2000) and the precentral gyrus, related to motor planning and execution (Schnitzler, Salenius, Salmelin, Jousmäki, & Hari, 1997). HDs had significantly greater activation Fed-Fasted in regions linked to reward and reinforcement (nucleus accumbens, substantia nigra and caudate; Benningfield et al., 2014; Nakanishi, Hikida, & Yawata, 2014) as well as the salience network of cingulate cortex (Seeley et al., 2007).

    • Effect of emotional valence on retrieval-related recapitulation of encoding activity in the ventral visual stream

      2015, Neuropsychologia
      Citation Excerpt :

      There is accumulating interest in the effect of emotion on the encoding-to-retrieval overlap (Dew et al., 2014; Hofstetter et al., 2012; Ritchey et al., 2013; Smith et al., 2004; Taylor et al., 1998) because the overlap of processes engaged during encoding and retrieval influences the probability of a memory coming to mind (Morris et al., 1977; Tulving and Thomson, 1973).

    • Bursts of transcranial electrical stimulation increase arousal in a continuous performance test

      2015, Neuropsychologia
      Citation Excerpt :

      Several studies on humans and animals have reported the key role of arousal in cognition (Berridge and Waterhouse, 2003; Brown et al., 2014). In humans, arousal is commonly modulated through emotional stimuli (Sutherland and Mather, 2012; Dew et al., 2014), warning cues (Hackley, 2009) and conflict paradigms (Brown et al., 2014), inducing an improvement of the behavioural performance as arousal increased. It has been demonstrated that an arousing stimulus can amplify the effect of saliency in short-term memory (Sutherland and Mather, 2012) and even enhance visual perception (Zeelenberg and Bocanegra, 2010) and memory retrieval (Mather and Sutherland, 2011).

    View all citing articles on Scopus
    View full text