The influence of emotional priming on the neural substrates of memory: A prospective fMRI study using portrait art stimuli

Abstract

Events coupled with an emotional context seem to be better retained than non-emotional events. The aim of our study was to investigate whether an emotional context could influence the neural substrates of memory associations with novel portrait art stimuli. In the current prospective fMRI study, we have investigated for one specific visual art form (modern artistic portraits with a high degree of abstraction) whether memory is influenced by priming with emotional facial pictures. In total forty healthy female volunteers in the same age range were recruited for the study. Twenty of these women participated in a prospective brain imaging memory paradigm and were asked to memorize a series of similar looking, but different portraits. After randomization, for twelve participants (Group 1), a third of the portraits was emotionally primed with approach-related pictures (smiling baby faces), a third with withdrawal-related pictures (baby faces with severe dermatological conditions), and another third with neutral images. Group 2 consisted of eight participants and they were not primed. Then, during an fMRI session 2 h later, these portraits were viewed in random order intermixed with a set of new (previously unseen) ones, and the participants had to decide for each portrait whether or not they had already been seen. In a separate experiment, a different sample of twenty healthy females (Group 3) rated their mood after being exposed to the same art stimuli, without priming. The portraits did not evoke significant mood changes by themselves, supporting their initial neutral emotional character (Group 3). The correct decision on whether the portraits were Familiar of Unfamiliar led to similar neuronal activations in brain areas implicated in visual and attention processing for both groups (Groups 1 and 2). In contrast, whereas primed participants showed significant higher neuronal activities in the left midline superior frontal cortex (Brodmann area (BA) 6), unprimed volunteers displayed higher right medial frontal cortical (BA 10) activities. Furthermore, specifically in Group 1, correct retrieval of negatively primed portraits evoked increased neuronal activity in the left medial orbitofrontal cortex (BA 11) and in the right (posterior) insula, suggesting enhanced stress-related responses to the memory of withdrawal-related primed modern artistic portraits in this group. Our prospective memory data in healthy females indicate that, to reach a correct retrieval decision, different midline anterior neuronal networks are recruited for portraits that were emotionally primed than for the unprimed ones. Importantly, our results also suggest that the negative emotional context leads to the formation of associations that are reactivated during memory retrieval processes of the initially neutral art portraits. When correctly recognized, the portraits evoke neuronal activities consistent with the withdrawal-related character of the emotional visual stimuli with which they have been associated. Although our results show that abstract portrait art can be associated with emotional primes this doesn't mean that this effect is specific for art images.

Introduction

Emotionally charged, novel, beautiful or aversive visual stimuli are better remembered than neutral or non-emotional ones (Wright et al., 2001). For instance, would Edward Munch's painting ‘The Scream’ leave the same indelible memory if the figure in the painting was not depicted with such an emotionally disturbing and agonizing face of terror (Solso, 2003)? Further, emotional events seem to be better retained than non-emotional events and emotional information is better remembered in mood states with a matching affective valence (Dolcos and Cabeza, 2002).

Nowadays, art and its creators have entered neuroscience research (Solso, 1994, Solso, 2003, Zeki, 1999). Images of art – paintings, sketches and photographs – have been used in a variety of brain imaging paradigms investigating their influence on perception, imagination, emotion and memory processes (Ishai et al., 2007, Lacey et al., 2011, Zaidel and Kasher, 1989). The few studies that examined neuronal activation in relation to visual art exposure suggest that perception and memory depend on visual as well as semantic brain processes (Cupchik et al., 2009, Fairhall and Ishai, 2008, Ishai et al., 2007). Also attention networks, such as the prefrontal and parietal cortices, and memory-related brain areas such as the temporal and hippocampal regions have been documented to be involved (Wiesmann and Ishai, 2008, Yago and Ishai, 2006), as well as reward-related regions such as the ventral striatum and the orbitofrontal cortices (Lacey et al., 2011).

How the brain stores and retrieves art-related memories seems to be a relevant research question given the complex interaction between the perception, emotional experience and memory processes involved. The emotional experience induced by art with a high degree of abstraction is not so easy to understand. Its valence can partly be determined by its previous association with an emotional context. Many affective priming studies examining memory processes, not always accompanied by brain imaging procedures, have reported behavioral effects predominantly after priming with negative stimuli (Chao and Yeh, 2008, Stolz and Neely, 2001, Zhang et al., 2006). Intriguingly, to our knowledge, no brain imaging studies have so far examined the influence of emotional priming on experiencing art. Consequently, in this paper we have examined this problem in the context of memory of portrait art with a high degree of abstraction.

For this study, we developed a new paradigm with novel stimuli based on modern paintings of faces. The portraits were selected from the work of a Belgian painter, Ronny Delrue (see Fig. 1B). This choice had several major advantages. Firstly, all participants were unfamiliar with the paintings, excluding recognition bias. Secondly, all paintings typically shared certain visual similarities, making it harder to remember them. Thirdly, the selected artworks have a high degree of abstraction and elicit no specific emotion by themselves. In addition, as the paintings were rendered in black and white, there was no ‘contamination’ by emotional connotations associated with different colors.

In our experiment, for one group of participants, a number of these portraits were primed by a preliminary viewing in combination with ‘emotional’ color pictures of baby faces. Among these there were positive or approach-related pictures showing happy, smiling baby faces. Negative or withdrawal-related pictures showed baby faces exhibiting a combination of negative facial expressions and severe dermatological conditions. Our earlier research in healthy women had already demonstrated that viewing the positively valenced baby pictures evoked strong approach-related emotions, such as happiness, and the negatively valenced ones strong withdrawal-related emotions, such as disgust (Baeken et al., 2010a). Furthermore, these emotional pictures had proven to be especially useful for examining brain activity related to emotional processes, in healthy as well as in psychopathological states (Baeken et al., 2009, Baeken et al., 2010a, Baeken et al., 2010b, Baeken et al., 2010c, Baeken et al., 2011).

As it had been demonstrated earlier that gender differences could affect emotional memory processes (Cahill et al., 2004, Piefke et al., 2005), and because age differences might affect memory paradigms (Rendell et al., 2011, Schnitzspahn et al., in press), we restricted our study cohort to healthy female subjects within a narrow age range. One group of healthy women (Group 1) was ‘primed’ with positive, neutral or negative baby faces while viewing a set of 60 Delrue portraits. A comparable sample of healthy women (Group 2) was not primed and served as the control group. A couple of hours later, these portraits, intermixed in random order with 60 new but similar ones, were viewed under event related fMRI. The viewing was combined with a forced recognition task: for each portrait participants had to choose between “I remember this image” and “I don't remember this image” by pressing the relevant buttons on two reaction boxes. Response accuracy and reaction times (RT) were recorded. In order to check that the portraits by themselves did not provoke mood alterations, in a separate study unrelated to the imaging study, a different but comparable female group rated their mood with Visual Analogue Scales (VAS; McCormack et al., 1988) before and after viewing all portraits (Group 3).

We expected that during the memory task all participants would primarily recruit brain areas related to visual processing and memory. Moreover, we hypothesized that emotionally primed portraits would, during memory retrieval, result in activation patterns within the fronto-limbic neurocircuitries consistent with the emotional valence of the prime. Finally, as most emotional priming studies reported ffects after priming with negative stimuli, we anticipated that these patterns would be primarily apparent for the negatively primed portraits.