Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2015 Jan 12.
Published in final edited form as: Neuroimage. 2011 Oct 1;59(2):1692–1699. doi: 10.1016/j.neuroimage.2011.09.029

BOLD responses in reward regions to hypothetical and imaginary monetary rewards

Krishna P Miyapuram a,*, Philippe N Tobler b,1, Lucy Gregorios-Pippas a, Wolfram Schultz a
PMCID: PMC4290830  EMSID: EMS61778  PMID: 21985912

Abstract

Monetary rewards are uniquely human. Because money is easy to quantify and present visually, it is the reward of choice for most fMRI studies, even though it cannot be handed over to participants inside the scanner. A typical fMRI study requires hundreds of trials and thus small amounts of monetary rewards per trial (e.g. 5p) if all trials are to be treated equally. However, small payoffs can have detrimental effects on performance due to their limited buying power. Hypothetical monetary rewards can overcome the limitations of smaller monetary rewards but it is less well known whether predictors of hypothetical rewards activate reward regions. In two experiments, visual stimuli were associated with hypothetical monetary rewards. In Experiment 1, we used stimuli predicting either visually presented or imagined hypothetical monetary rewards, together with non-rewarding control pictures. Activations to reward predictive stimuli occurred in reward regions, namely the medial orbitofrontal cortex and midbrain. In Experiment 2, we parametrically varied the amount of visually presented hypothetical monetary reward keeping constant the amount of actually received reward. Graded activation in midbrain was observed to stimuli predicting increasing hypothetical rewards. The results demonstrate the efficacy of using hypothetical monetary rewards in fMRI studies.

Keywords: Reward, Hypothetical reward, Money, fMRI, Dopaminergic, Midbrain, medial orbitofrontal cortex, ventral striatum, Classical conditioning, Reward imagination, Mental imagery, Neuroeconomics

Introduction

Money is a strong motivator. Subjects work harder, more persistently, and more effectively, if they earn more money for better performance (Camerer and Hogarth, 1999). Typical fMRI studies involve hundreds of trials with each trial lasting around 10 s. If one wants to treat each trial equally and incentive-compatible, then monetary rewards paid to the participants per trial need to be small to avoid excessive expenditure for the experimenter. Gneezy and Rustichini (2000) found that while larger amounts of money yielded higher performance (in terms of parents complying with day care rules), smaller amounts yielded poorer performance than no compensation at all. Hence, monetary incentives at low value can have a detrimental effect on performance. Hypothetical payoffs provide one possibility to overcome the limitation of small gambles in experimental situations (Kahneman and Tversky, 1979).

Hypothetical rewards have been investigated in behavioral studies using paradigms such as the Iowa gambling task (Bowman and Turnbull, 2003; Fernie and Tunney, 2006) and temporal discounting tasks (Johnson and Bickel, 2002; Lagorio and Madden, 2005). These previous studies suggest that hypothetical rewards can motivate behavior to a similar degree as actual rewards (but see Holt and Laury, 2002). In a neuroimaging study, Wittmann et al. (2007) investigated temporal discounting using hypothetical rewards in the range of few hundred dollars with delays ranging from 5 days to 10 years. However, hypothetical rewards have also been used for much shorter delays (e.g. in Gregorios-Pippas et al., 2009, where participants received a percentage of displayed reward every trial). Neural findings from these earlier studies using hypothetical rewards were in line with other imaging studies of temporal discounting using potentially real rewards (e.g. McClure et al., 2004, where one of the participant’s choices was selected randomly). In agreement with this notion, Bickel et al. (2009) found no difference in behavioral and neural activation when comparing real and hypothetical monetary gains and fictive monetary losses in a temporal discounting task. Thus, at least in situations involving future outcomes, hypothetical monetary rewards seem to be able to influence behavior and neural activity similar to real payoffs.

Monetary rewards are secondary reinforcers that possibly acquired their value by association with more primary rewards. It is therefore conceivable that monetary rewards act through imagination of rewarding objects that can be acquired through them. Indeed in a recent reward imagination study by Bray et al. (2010), participants who imagined monetary or other types of secondary rewards activated primary reward regions. However, monetary rewards have become so ubiquitous in our daily life that they can be considered as rewards in their own right. Functional neuroimaging studies using monetary rewards have found the same brain regions as those involved in processing primary rewards also without asking participants to imagine (e.g. Kim et al., 2011; Tobler et al., 2007; Valentin and O’Doherty, 2009). Given that money is easier to present than primary rewards, it is no surprise that monetary rewards have become the choice of most neuroimaging studies of reward processing. We set out to investigate motivational aspects and corresponding neural activity using various different ways of presenting hypothetical monetary rewards.

Human neuroimaging studies using blood-oxygen-level-dependent (BOLD) functional magnetic resonance imaging (fMRI) have identified a set of reward regions in the midbrain, striatum and orbitofrontal cortex, consistent with findings from non-human primates (for review see Haber and Knutson, 2010). Moreover, conditioned stimuli predicting real monetary rewards activate reward regions in humans (e.g. Tobler et al., 2007). Here, we set out to investigate whether also predictors of hypothetical monetary rewards would activate reward regions of the brain. In order to investigate motivational aspects of monetary rewards, we presented participants with hypothetical rewards that displayed a larger value, but told them that they would be able to take home only a specific percentage of the displayed money. Some previous studies (e.g. Dreher et al., 2006; Gregorios-Pippas et al., 2009; Kuhnen and Knutson, 2005) have used a similar strategy but not explicitly studied the effects of hypothetical rewards. Accordingly, we performed two experiments using different scenarios – when the hypothetical rewards are visually presented or imagined (Experiment 1) and when the real pay-off is explicitly displayed (Experiment 2) or invisible (Experiment 1). Furthermore we hypothesized that increasing the hypothetical monetary reward would result in increased activation in reward regions although the real pay-off is kept constant (Experiment 2).

Materials and methods

Subjects

A total of 17 right handed participants (mean age 25.5 years, range 21.6–30.5 years) were recruited from among undergraduate and graduate student populations in the Cambridge area of the UK. Ten participants took part in both the experiments (the first experiment was always performed at the end of the session). The data from two participants could not be recovered due to technical reasons. The data from three participants, from each of the two experiments, were excluded from all analyses due to large head movements detected in the realignment process. The final data analysis was performed on 12 participants each in the first (5 female) and second (6 female) experiments. Participants were pre-assessed through self-report to exclude prior histories of neurological or psychiatric illness (including depression and schizophrenia), and drug usage. Participants were asked to refrain from excessive alcohol usage on the night before scanning. All participants gave informed written consent and were paid for participation. The research protocol was approved by the Cambridgeshire Local Research Ethics Committee, U.K.

Stimuli, training and pleasantness ratings

Scanning only took place after participants had learned the meaning of abstract visual stimuli in a training session (see below). Experiment 1 used a 2×2 factorial design in which four abstract stimuli were paired with visually presented or imagined hypothetical monetary reward or a control picture. During the imagination trials, a blank screen was displayed following the offset of the abstract stimulus. Experiment 2 used a 2×3 factorial design in which six abstract stimuli were paired with pictorial or alphanumerical display of parametrically increasing hypothetical monetary rewards with three different magnitudes. The hypothetical monetary rewards were always accompanied by a fixed real-payoff (5 pence) that was invisible (Experiment 1) or explicitly displayed (Experiment 2). The stimulus–reward associations were randomized across subjects for both experiments. The presentation of the stimuli and recording of responses was controlled by Cogent 2000 software on Matlab.

For both experiments 1 and 2, we trained subjects up to a week before the scanning session to learn the correct stimulus–reward contingencies (Fig. 1a). Stimulus pleasantness ratings for the abstract stimuli were collected from the subjects before and after training and scanning sessions (Fig. 1a). The abstract stimuli were rated on a scale of 1 (dislike very much) to 5 (like very much). The value of 3 was marked as indifferent on the pleasantness rating scale. After baseline pleasantness ratings were collected, participants received a customized instruction sheet, showing the exact stimulus–reward contingencies. After reading the instructions, participants returned the instruction sheet to the experimenter. The training session proceeded with identical trial structure (see descriptions below) as the scanning session except that the number of repetitions for each trial type in the training session was half (Experiment 1: 15, Experiment 2: 20) of those in the scanning session.

Fig. 1.

Fig. 1

Open in a new tab

Experimental design and behavioral results. (a) Training and scanning schedule. Participants were trained up to one week before the scanning session using exactly the same stimulus–reward contingencies. In the first experiment, the instructions to imagine were given in a customized instruction sheet showing the stimulus–reward contingencies for every participant. They reported after every trial what they had seen or imagined using a button press. Participants performed an additional testing task immediately after the training of the second experiment. The training and scanning sessions for the two experiments were identical except that the training sessions had only half the number of trials. Pleasantness ratings for abstract stimuli were collected at the beginning and end of training and scanning sessions. The training and scanning sessions took place on different days separated up to one week. The first experiment was always at the end of the session for those participants who took part in both experiments. (b) Experiment 1. Four conditional stimuli were associated with either rewarding or non-rewarding pictures visually presented on the screen or to be imagined during a blank screen. Participants were instructed before the start of the experiment to imagine the corresponding reward or non-rewarding picture when the blank screen appeared. To ensure minimal demand effects, an invisible actual reward was given for every rewarding trial. The invisible reward was by two orders of magnitude smaller than the hypothetical reward picture and participants were fully informed of this manipulation at the start of the experiment. The contents of visual presentation and imagination were recorded by self-report on every trial. (c) Experiment 2. Three conditional stimuli were associated with three hypothetical rewards that yielded the same actual reward. Two modes of reward presentation were used for hypothetical rewards – alphanumeric and pictorial. The actual reward was not displayed to the participants but added to the total take-home sum that was continuously displayed at the bottom of the screen throughout the experiment.

Experiment 1

In Experiment 1 (Fig. 1b), four visual (fractal) stimuli predicted either visual presentation or imagination of a money bill (reward) or a scrambled picture (no reward). The rewarding picture was a hypothetical monetary reward (£20 money bill) together with an invisible real payoff (5 pence) that was added to the cumulative sum to take home at the end of the experiment. The non-rewarding picture was a scrambled picture of a £20 money bill that did not contribute towards cumulative earnings. The scrambled picture was smoothed to avoid any overt association with the £20 money bill or reconstruction of the money bill from its parts. The visual presentation and imagination trials differed in that the money bill (hypothetical reward) and the scrambled picture were displayed on the screen or not, respectively.

Each trial started with a central fixation ‘+’ symbol at the centre of the screen, presented during a truncated Poisson distributed random interval (mean 4 s; maximum duration 6 s). Then, one of the four visual stimuli appeared for 2 s. The offset of the visual stimulus marked the onset of a 1 s presentation of the rewarding picture (£20 money bill) or the scrambled picture for visual presentation trials or of a blank screen for imagination trials. At the end of every trial, one of the two questions “what did you see” or “what did you imagine” appeared on the screen. Below the question, three answer choices (money bill; scrambled picture; nothing) were indicated. Participants responded by pressing one of three corresponding buttons. The spatial position of the three answer choices on the screen was randomized across trials to avoid preparation of specific movements during stimulus presentation. Including the option of having seen or imagined “nothing” allowed us to get reliable self-report from the participants. If the participant did not respond within 1.5 s, a red square was flashed for 0.5 s and the trial was repeated. Cumulative earnings were not displayed to prevent explicit pairing of the visual stimuli with the visual receipt of the reward in the imagination trials. Unbeknown to the participants, a total of 30 repetitions for each trial type were required to complete the scanning session.

In this experiment, we particularly ensured that participants successfully understand the meaning of stimuli that predicted imagination of the money bill or the scrambled picture by the customized instruction sheet. No further explicit instructions were given during the experiment regarding imagination. Hence, participants never experienced a visible (hypothetical or real) reward in the imagination trials in either the training or scanning sessions. We adopted this strategy so that the training and scanning sessions were identical. Learning was ensured by including a question at the end of every trial (as indicated above).

Experiment 2

Experiment 2 investigated the effects of parametrically increasing hypothetical rewards without increasing the actual pay-off (Fig. 1c). Six abstract stimuli were associated with three values of alphanumeric or pictorial displayed money – 10 pence, 1 Pound and 10 Pounds, all predicting the same amount of real money (5 pence). Participants were informed before the experiment that all displayed money was hypothetical whereas actual cumulative earnings were displayed at the bottom of the screen throughout the experiment. To control for visual saliency, we presented both pictures of real money and alphanumeric representations of the three values. Monetary rewards were visually presented at a uniform size of 240×120 pixels.

Each trial started with a central fixation ‘+’ symbol at the centre of the screen, presented during a random interval that followed a truncated Poisson distribution (mean 5 s; maximum duration 8 s). This interval served to separate the present from the previous trial. Then, one of the six visual stimuli (dimensions 100×100 pixels) appeared for 2 s at a random location either on the top-half or on the bottom-half of the screen. A conditional motor task was overlaid onto the Pavlovian paradigm to keep participants attentive throughout the experiment. Participants responded by pressing either the left or right button (using a mouse in training sessions and an MRI compatible button box in scanning sessions) with their index and middle fingers, respectively, to indicate where on the screen the stimulus had appeared. The allocation of top half-to-left button vs. top half-to-right button was randomized across participants. After a correct response, the monetary reward was visually presented for 1 s replacing the abstract stimulus on the screen. Upon error, a red square was shown at the centre of the screen and the trial was repeated. The cumulative sum of actual money received was displayed throughout at the bottom of the screen, and was updated at the time of displaying the monetary reward. A total of 40 repetitions for each trial type were required to complete two scanning runs. The counter of cumulative earnings was reset in-between two scanning runs.

To ensure learning of stimulus–reward contingencies, participants were tested on a modified version of the task immediately after the training session. The testing task consisted of displaying one abstract stimulus at a time at the centre top half of the screen and the six sets of monetary rewards in two rows at the bottom of the screen. Subjects indicated the monetary reward associated with the abstract stimulus displayed by pressing the appropriate key on a keypad.

Data acquisition and analysis

This section describes imaging data acquisition and analysis for both experiments. Functional imaging was performed on a MedSpec (Bruker, Ettlingen, Germany) scanner at 3 T. Two whole brain acquisitions weighted by Gradient Echo and Spin Echo were obtained within a scan repetition time (TR) of 2.4 s. 21 horizontal slices with in-plane resolution of 3.75 mm×3.75 mm, were collected with a slice thickness of 5 mm and slice gap of 1 mm. We used Statistical Parametric Mapping (SPM2) software from the Wellcome Department of Cognitive Neurology, London within Matlab 6.5 (The Mathworks Inc., MA) for data analysis. The first five volumes of each series were discarded to avoid T1 saturation effects. Results from the Gradient Echo acquisition are reported as no additional activation was observed with the Spin Echo acquisition.

Images were slice-time corrected and realigned to correct for any head movements. The realigned images were normalized to a standard brain EPI template in MNI (Montreal Neurological Institute) space (Friston et al., 1995a) and resampled to a voxel size of 2×2×2 mm. An isotropic Gaussian filter with 8 mm full width at half maximum was used to spatially smooth the normalized images. Serial autocorrelations within the functional data were estimated using a first order auto-regressive model (AR-1). Statistical analysis was performed using a general linear model (Friston et al., 1995b). The data were high-pass filtered at a frequency of 1/128 Hz. Given that the stimulus–outcome associations had already been learnt during the training session by our participants, activations were assessed at the onset of the abstract visual stimuli predicting reward or no reward (Experiment 1) or hypothetical rewards of different magnitudes (Experiment 2). Stick functions at the onset of the visual stimuli were convolved with a canonical hemodynamic response function (HRF) and its time and dispersion derivatives. Group analysis was performed with random effects models (Penny et al., 2003) and correction for non-sphericity of the data.

Thresholding strategy

Based on previous neurophysiological and human neuroimaging studies of reward and novelty, we defined three regions of interest (ROIs) in the midbrain (Bunzeck and Düzel, 2006; Elliott et al., 2003; O’Doherty et al., 2006; Schultz, 1998; Tobler et al., 2007), ventral striatum (Burke et al., 2010; Knutson et al., 2001; O’Doherty et al., 2006) and medial orbitofrontal cortex (Bray et al., 2010; O’Doherty et al., 2001; Sescousse et al., 2010) in agreement with their known role in reward processing. All ROIs were anatomically defined using the WFU Pick Atlas toolbox (Lancaster et al., 2000; Maldjian et al., 2003). The ROI for the midbrain consisted of bilateral substantial nigra extended by 2 mm to include the medially adjoining ventral tegmental area (VTA). Given the similarity of neurophysiological responses of dopamine neurons in the ventral tegmental area and the substantia nigra pars compacta (Schultz, 1998), we did not focus on either of these structures (D’Ardenne et al., 2008; Düzel et al., 2009). The ROI for ventral striatum was anatomically defined consisting of caudate and putamen confined to brain slices z<0 mm and extended by 2 mm to include the medially adjoining nucleus accumbens. The ROI for medial OFC was anatomically defined. Reported activations are corrected for multiple comparisons (p<0.05, family-wise error) for height of activation within the volume of our ROI. For visualization of the extent of activations in our ROIs (in figures and activation table), we use an exploratory threshold of p<0.005 uncorrected with a minimum extent of 10 voxels.

Results

Behavioral results

Pleasantness ratings of abstract stimuli did not differ before training (Experiment 1: F(3,33)=0.49, p=0.69; Experiment 2: F(5,66)=0.59, p=0.7). Participants successfully learned the stimulus–reward associations by the end of the training session for both experiments (Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Pleasantness ratings. (a) Pleasantness ratings for four conditional stimuli in Experiment 1 on a scale of 1 (dislike very much) to 5 (like very much). Pretraining took place up to one week before the scanning session. (b) Pleasantness ratings for the six conditional stimuli predicting three hypothetical rewards in alphanumeric (left panel) and pictorial (right panel) modes of presentation.

In Experiment 1, the stimuli predicting reward were rated as more pleasant than the control stimuli predicting no reward (visual presentation: p<0.005, imagination: p<0.05). In Experiment 2, stimuli associated with 10 pounds were rated as more pleasant than stimuli associated with 10 pence (alphanumeric: p<0.05, pictorial: p<0.05). Learning, as indexed by change in pleasantness ratings, was significant during the training session (Experiment 1: p<0.05, Experiment 2: p<0.05). No further learning took place during the scanning session (first experiment: p=0.07, second experiment: p=0.12). These results suggest that the stimulus–reward associations were successfully established during the training session, which took place up to a week before the scanning session.

In Experiment 1, all participants correctly reported the contents of visual presentation and imagination (money bill or scrambled picture) at the end of every trial in more than 94% of trials (average 96.8%±1%) in the scanning session. In Experiment 2, participants correctly identified stimulus–reward associations (accuracy ranging between 83.6±7% and 93.3±3.5%) during the testing session immediately after the training session. These results indicate the correct identification of stimulus–reward associations.

Brain imaging

Activation to hypothetical and invisible rewards

The main effect of brain activation to hypothetical rewards is identified from Experiment 1 by contrasting stimuli predicting reward with stimuli predicting no reward. The reward in this comparison consisted of visually presented or imagined hypothetical monetary reward (20 pounds money bill) together with an invisible real pay-off (5 pence). We found a main effect for reward-related activation in the medial orbitofrontal cortex (OFC) and the midbrain (Fig. 3, Table 1). Activation in the ventral striatum did not survive small volume correction but was present at a lower threshold (Fig. 3, Table 1). When tested individually, reward-related activation was observed in the medial OFC and the midbrain in visual presentation and imagination trials, respectively (Fig. 4, Table 1). However, direct comparison of visual presentation and imagination trials did not reveal differential activations in any of the ROIs even at the exploratory threshold (p<0.005, k=10 voxels). At this exploratory threshold, we found overlapping, common activation in the midbrain (p=0.07, small volume corrected) in the imagination and visual presentation trials using conjunction analysis (Nichols et al., 2005). These results indicate that hypothetical monetary rewards (irrespective of whether they are visually presented or imagined) enter associations with predictive stimuli and thereby activate reward regions even though the real pay-off is small.

Fig. 3.

Fig. 3

Open in a new tab

Main effect of reward-related activation. Activation in (a) medial orbitofrontal cortex, (b) midbrain and (c) ventral striatum for reward compared to no reward in Experiment 1. The activation maps are thresholded at p<0.005 and an extent threshold of 10 voxels for display purposes. Note that activations are displayed for the whole brain and the anatomical ROI masks are not used in the figures.

Table 1.

Location of peak activations in regions of interest (ROIs). Extent of activation (cluster size) is reported at uncorrected threshold of p<0.005 and the number of voxels surviving small volume correction for height of activation is also given.

Contrast ROI MNI coordinates (mm)
 Z score Cluster size
x y z Uncorrected ROI corrected
Experiment 1
Main effect: reward>no reward Medial OFC 10 46 −8 4.14 254 32
Midbrain 12 −18 −6 3.46 56 3
−6 −8 −12 3.13 68 0
−8 −18 −6 3.12
Ventral striatum 6 10 −10 2.85 37 0
Visual presentation (reward>no reward) Medial OFC 10 46 −8 4.17 246 35
Midbrain 10 −20 −6 3.10 12 0
Ventral striatum 2 12 −12 3.44 80 0
Imagination (reward>no reward) Midbrain −6 −10 −6 3.54 80 8
8 −20 −6 3.46 43 1
Conjunction: visual presentation and imagination (rewardNno reward) Midbrain 10 −20 −6 3.10 11 0
Experiment 2
Main effect: 10 pounds>10 pence Midbrain 10 −10 −12 3.52 55 4
Alphanumeric (10 pounds>10 pence) Midbrain 4 −16 −8 2.92 22 0
−4 −16 −10 2.81 19 0
Pictorial (10 pounds>10 pence) Midbrain 8 −26 −12 2.80 4 0
Open in a new tab
Fig. 4.

Fig. 4

Open in a new tab

Activation to visually presented and imagined rewards. (a, b) Medial orbitofrontal cortex (OFC), (c–e) midbrain. Activation is shown in green for visual presentation condition and in red for imagination condition. See the legend of Fig. 3 for thresholding information. The corresponding contrasts of parameter estimates (betas) are plotted in (b) medial OFC, (d), left and (e) right midbrain.

Graded activation to hypothetical rewards

The second experiment investigated the effect of parametrically increasing the value of the visually presented hypothetical reward (10 pence, 1 pound, 10 pounds) without increasing the real pay-off (5 pence). Two modes of hypothetical rewards were used – pictures of real money or alphanumeric displays of the corresponding value. By displaying the real cumulative take-home amount of money, we ensured that the brain responses reflect the effect of increasing the hypothetical monetary rewards. A linear contrast compared the stimuli predicting 10 pounds to stimuli predicting 10 pence. The contrast weight for reward display of 1 pound was set to zero, thus served as a middle point. Graded activation in the midbrain was observed with this contrast irrespective of whether the hypothetical rewards were alphanumeric or pictorial (Fig. 5, Table 1). We did not find activations in other ROIs even at the exploratory threshold. The opposite contrast comparing 10 pence>10 pounds with 1 pound as middle point did not yield any activations in our ROIs even at an exploratory threshold. We also tested an ROI in the striatum that included the dorsal striatum and did not find any activation for the contrast of interest. These results indicate that the midbrain activation reflects increasing magnitude of hypothetical monetary rewards although the real pay-off is kept constant.

Fig. 5.

Fig. 5

Open in a new tab

Graded activation in the midbrain to hypothetical rewards in Experiment 2 and the corresponding plot of parameter estimates (betas). See also the legend of Fig. 3 for thresholding information.

Discussion

This research investigated brain activations to hypothetical monetary rewards that displayed an increased value of monetary rewards without in fact increasing the actual reward paid to the subject. The results show activations in the medial orbitofrontal cortex and the midbrain in processing visually presented and imagined hypothetical monetary rewards. Furthermore, midbrain activation increased in a graded fashion with increasing amounts of hypothetical monetary rewards. Our results suggest that the use of hypothetical monetary rewards could be adopted as a generic technique for performing neuroimaging studies on reward.

Monetary rewards in neuroimaging studies

Functional neuroimaging studies using monetary rewards have found the same brain regions as those involved in processing primary rewards (e.g. Kim et al., 2011; Tobler et al., 2007; Valentin and O’Doherty, 2009). These and other previous studies have used varied experimental procedures to determine the way subjects receive monetary rewards. For instance, a consolidated amount is paid for participation and the reward is not contingent on the subject’s performance (e.g. Boettiger et al., 2007; Schonberg et al., 2010). Another approach would be to use both monetary gains and losses (e.g. Delgado et al., 2005; Kim et al., 2006) so that the total money payable to subjects at the end of the experiment is not very large. Subjects might prefer (infrequent) large rewards compared to cumulative (frequent) small rewards. Hence, some paradigms use probabilistic rewards or randomly pick one of the decisions made about large rewards by the subject (e.g. Hsu et al., 2005; Pine et al., 2009).

Hypothetical monetary rewards (see also Bickel et al., 2009; Dreher et al., 2006; Kuhnen and Knutson, 2005) offer a good alternative to these methods of payment to subjects, as it offers stringent control to experimenters without having to resort to deceiving subjects, using probabilistic outcomes or losses, or treating trials differently. An unavoidable limitation of the approach used in our experiments is that the magnitude of the displayed reward covaried with the hypothetical monetary reward. To account for this, participants were explicitly informed about the actual reward they would receive in each trial. Using hypothetical rewards also overcomes the issue that the real value of monetary rewards given every trial (e.g. 5p) has to be very small. It brings together psychological theories that postulate a positive reinforcement such as an OK signal is sufficient to drive behavior, and economic theories that suggest monetary reward is crucial to positively drive behavior (Camerer and Hogarth, 1999), as long as the monetary reward is not too small (Gneezy and Rustichini, 2000). As an extension of our approach, it would be interesting to compare the receipt of positive reinforcement signals and other forms of hypothetical rewards such as receiving ‘points’.

Reward regions activated by hypothetical monetary rewards

Experiment 2 revealed graded activations in the midbrain to stimuli predicting different reward denominations (Fig. 5, Table 1). The design of this study allowed us to dissociate the visual aspects of the monetary rewards from the value attached to them by using both alphanumeric and pictorial representations. The activation observed in the midbrain occurred in a region that also responds to primary liquid and food reward (Schultz, 1998). Similar activations occur in midbrain neurons of humans with monetary rewards (Zaghloul et al., 2009).

The graded activations in midbrain to increasing hypothetical rewards occurred despite the fact that each trial resulted in the same real pay-out. One possibility to account for this finding is that the reinforcing properties of money are so strong that when tested under extinction (such as delivering 5 pence when displaying 5 pounds, thus artificially decreasing the value of displayed reward), the responses do not extinguish. In other words, the graded responses comply with the displayed reward magnitude, instead of diminished responses to the extinction procedure. Such resistance to extinction would be reminiscent of the effects of drugs of abuse on instrumental behavior (e.g. Quick and Shahan, 2009).

Experiment 1 revealed activation in medial OFC to visually presented hypothetical monetary rewards (Fig. 4, Table 1). Previous neuroimaging studies have found a role for OFC in outcome valuation (Chib et al., 2009; Elliott et al., 2008; Peters and Büchel, 2009, 2010; Sescousse et al., 2010) and a medio-lateral distinction in processing rewarding and aversive stimuli (Elliott et al., 2000; Kringelbach and Rolls, 2004; O’Doherty et al., 2001). The activation in the medial OFC to stimuli predicting visually presented rewards is consistent with these earlier findings. However, at the employed statistical thresholds we did not find activations in the medial OFC ROI to increasing hypothetical reward. The outcomes in terms of real pay-off were the same irrespective of the displayed reward. Hence, it could be the case that the medial OFC is not differentially engaged when the actual outcomes do not differ.

In our two experiments, we did not observe reliable activations in the ventral striatum at the employed thresholds. Experiment 1 revealed a main effect for reward in the ventral striatum at an exploratory threshold (Fig. 3, Table 1). At this threshold, activation in ventral striatum was observed in visual presentation trials (Table 1). In Experiment 2, we did not find increasing activation in ventral striatum with increasing hypothetical rewards. Earlier neuroimaging studies have implicated the ventral striatum in representing reward value (Knutson et al., 2001; Peters and Büchel, 2009, 2010) and in reward prediction error (Hare et al., 2008; McClure et al., 2003; Pagnoni et al., 2002; Tobler et al., 2006). In a temporal discounting paradigm with similar hypothetical monetary rewards as in our experiments, ventral striatum activation showed differential discounting depending on the magnitude of the reward (Gregorios-Pippas et al., 2009). The real pay-offs in Gregorios-Pippas et al. (2009) were proportional to the displayed reward, whereas in our experiment, the real pay-off was constant irrespective of the displayed reward. Hence, it could be the case that, similar to medial OFC, the ventral striatum is not differentially engaged when the actual outcomes do not differ.

Together, the absence of graded responses in medial OFC and ventral striatum (see Peters and Büchel, 2010) suggests that hypothetical rewards could be less efficient in eliciting increasing activations with increasing reward magnitude. A direct comparison of neural responses between increasing monetary reward and increasing hypothetical reward would be an interesting extension of our study.

Imagination of hypothetical monetary rewards

A novel contribution from our study is the investigation of experienced and imagined rewards. We found activation in medial OFC and midbrain to stimuli predicting visually presented and imagined rewards, respectively, though these activations were not significantly different among the two conditions (Fig. 4, Table 1). Bray et al. (2010) found activation in medial OFC when participants imagined items they found personally rewarding. In the present study, we found reliable midbrain activation extending bilaterally in the imagination trials. Our findings suggest a possibility of common neural activation in the midbrain for imagined and (albeit slightly less strong in) visually presented rewards. It is of interest that no other activation was found to be common outside this ROI in a whole brain exploratory analysis. One reason for the discrepancy between our findings and those of Bray et al. (2010) could be due to the short duration of reward in the current study that limited the generation of vivid mental images as opposed to the longer duration used in the block design of Bray et al. (2010).

A key limitation of our study is that we cannot dissociate common activation, if any, between visually presented and imagined trials to be due to the imagination of hypothetical reward or the common receipt of invisible reward (i.e. 5 pence). Our findings also have to be interpreted in view of the limited power provided by the relatively small number of participants (n=12). Including more participants should be a consideration for future studies according to current practices in neuroimaging. Furthermore, the image acquisition was done at a lower spatial resolution of 3.75×3.75×5 mm3. Higher resolution of functional imaging would be desirable when focusing of regions such as the midbrain (D’Ardenne et al., 2008).

While our data are somewhat suggestive about common neural activations for reward-related visual imagery and perception, Kosslyn et al. (1996) did not find common brain areas affected by aversive stimuli during imagery and perception in their positron emission tomography (PET) study. Bray et al. (2010) have demonstrated a role for medial OFC in reward imagination. Contrary to those studies, we have used an event-related design with randomized trials of reward imagery and perception. The difference between these diverse findings regarding common neural activity may point to subtle differences in the imagination of outcomes of different valence, although the differences could also be due to methodological differences, such as the higher temporal resolution of fMRI compared to PET. Further studies using both rewarding and aversive stimuli with the same methods in the same participants would be required to corroborate this potential difference.

Theories of mental imagery disagree on whether the content of mental images and percepts is visual depictions (Kosslyn, 1988) or abstract propositions (Pylyshyn, 2002). We verbally instructed participants at the outset that all stimuli associated with reward presentation and imagination would result in the same amount of reward at the end of the experiment. Hence, participants may have imagined verbal rather than visual contents. To more fully address the mental contents of reward imagination, future research may want to dissociate propositional and visual reward imagination more explicitly. As a further extension, it would also be interesting to compare imagination of abstract reinforcers such as money to more primary rewards such as food reward or erotic stimuli (e.g. Sescousse et al., 2010; Simmons et al., 2005).

In summary, we propose that despite their limitations, hypothetical monetary rewards can be used to study the human reward system in a straightforward manner. This technique is particularly useful to show participants monetary rewards with some buying power so that reward responses are maintained and avoid disappointment or frustration induced by the small value of money otherwise given at single trial levels.

Acknowledgments

This research was funded by the Wellcome Trust and the MRCWellcome Behavioural and Clinical Neuroscience Institute (BCNI). KPM was supported by the Cambridge Nehru Scholarship and Overseas Research Students Award Scheme, UK, and PNT by a Royal Society University Research Fellowship and the Swiss National Science Foundation (PP00P1_128574).

References

  1. Bickel WK, Pitcock JA, Yi R, Angtuaco EJ. Congruence of BOLD response across intertemporal choice conditions: fictive and real money gains and losses. J. Neurosci. 2009;29(27):8839–8846. doi: 10.1523/JNEUROSCI.5319-08.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Boettiger CA, Mitchell JM, Tavares VC, Robertson M, Joslyn G, D’Esposito M, Fields HL. Immediate reward bias in humans: fronto-parietal networks and a role for the catechol-O-methyltransferase 158(Val/Val) genotype. J. Neurosci. 2007;27(52):14383–14391. doi: 10.1523/JNEUROSCI.2551-07.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bowman C, Turnbull O. Real versus facsimile reinforcers on the Iowa Gambling Task. Brain Cogn. 2003;53:207–210. doi: 10.1016/s0278-2626(03)00111-8. [DOI] [PubMed] [Google Scholar]
  4. Bray S, Shimojo S, O’Doherty JP. Human medial orbitofrontal cortex is recruited during experience of imagined and real rewards. J. Neurophysiol. 2010;103(5):2506–2512. doi: 10.1152/jn.01030.2009. [DOI] [PubMed] [Google Scholar]
  5. Bunzeck N, Düzel E. Absolute coding of stimulus novelty in the human substantia nigra/VTA. Neuron. 2006;51:369–379. doi: 10.1016/j.neuron.2006.06.021. [DOI] [PubMed] [Google Scholar]
  6. Burke CJ, Tobler PN, Schultz W, Baddeley M. Striatal BOLD response reflects the impact of herd information on financial decisions. Front. Hum. Neurosci. 2010;4:48. doi: 10.3389/fnhum.2010.00048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Camerer CF, Hogarth RM. The effects of financial incentives in experiments: a review and capital-labor-production framework. J Risk Uncertainty. 1999;19:7–42. [Google Scholar]
  8. Chib VS, Rangel A, Shimojo S, O’Doherty JP. Evidence for a common representation of decision values for dissimilar goods in human ventromedial prefrontal cortex. J. Neurosci. 2009;29(39):12315–12320. doi: 10.1523/JNEUROSCI.2575-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. D’Ardenne K, McClure SM, Nystrom LE, Cohen JD. BOLD responses reflecting dopaminergic signals in the human ventral tegmental area. Science. 2008;319:1264–1267. doi: 10.1126/science.1150605. [DOI] [PubMed] [Google Scholar]
  10. Delgado MR, Miller MM, Inati S, Phelps EA. An fMRI study of reward-related probability learning. NeuroImage. 2005;24(3):862–873. doi: 10.1016/j.neuroimage.2004.10.002. [DOI] [PubMed] [Google Scholar]
  11. Dreher J, Kohn P, Berman K. Neural coding of distinct statistical properties of reward information in humans. Cereb. Cortex. 2006;16:561–573. doi: 10.1093/cercor/bhj004. [DOI] [PubMed] [Google Scholar]
  12. Düzel E, Bunzeck N, Guitart-Masip M, Wittmann B, Schott BH, Tobler PN. Functional imaging of the human dopaminergic midbrain. Trends Neurosci. 2009;32:321–328. doi: 10.1016/j.tins.2009.02.005. [DOI] [PubMed] [Google Scholar]
  13. Elliott R, Dolan RJ, Frith CD. Dissociable functions in the medial and lateral orbitofrontal cortex: evidence from human neuroimaging studies. Cereb. Cortex. 2000;10(3):308–317. doi: 10.1093/cercor/10.3.308. [DOI] [PubMed] [Google Scholar]
  14. Elliott R, Newman J, Longe O, Deakin J. Differential response patterns in the striatum and orbitofrontal cortex to financial reward in humans: a parametric functional magnetic resonance imaging study. J. Neurosci. 2003;23:303–307. doi: 10.1523/JNEUROSCI.23-01-00303.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Elliott R, Agnew Z, Deakin JF. Medial orbitofrontal cortex codes relative rather than absolute value of financial rewards in humans. Eur. J. Neurosci. 2008;27(9):2213–2218. doi: 10.1111/j.1460-9568.2008.06202.x. [DOI] [PubMed] [Google Scholar]
  16. Fernie G, Tunney R. Some decks are better than others: the effect of reinforcer type and task instructions on learning in the Iowa Gambling Task. Brain Cogn. 2006;60:94–102. doi: 10.1016/j.bandc.2005.09.011. [DOI] [PubMed] [Google Scholar]
  17. Friston KJ, Ashburner J, Frith CD, Poline JB, Heather JD, Frackowiak RSJ. Spatial registration and normalisation of images. Hum. Brain Mapp. 1995a;2:165–189. [Google Scholar]
  18. Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith CD, Frackowiak RSJ. Statistical parametric maps in functional imaging: a general linear approach. Hum. Brain Mapp. 1995b;2:189–210. [Google Scholar]
  19. Gneezy U, Rustichini A. Pay enough or don’t pay at all. Q. J. Econ. 2000;115:791–810. [Google Scholar]
  20. Gregorios-Pippas L, Tobler PN, Schultz W. Short-term temporal discounting of reward value in human ventral striatum. J. Neurophysiol. 2009;101(3):1507–1523. doi: 10.1152/jn.90730.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Haber SN, Knutson B. The reward circuit: linking primate anatomy and human imaging. Neuropsychopharmacology. 2010;35(1):4–26. doi: 10.1038/npp.2009.129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hare TA, O’Doherty J, Camerer CF, Schultz W, Rangel A. Dissociating the role of the orbitofrontal cortex and the striatum in the computation of goal values and prediction errors. J. Neurosci. 2008;28(22):5623–5630. doi: 10.1523/JNEUROSCI.1309-08.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Holt CA, Laury SK. Risk aversion and incentive effects. Am. Econ. Rev. 2002;92:1644–1655. [Google Scholar]
  24. Hsu M, Bhatt M, Adolphs R, Tranel D, Camerer CF. Neural systems responding to degrees of uncertainty in human decision-making. Science. 2005;310(5754):1680–1683. doi: 10.1126/science.1115327. [DOI] [PubMed] [Google Scholar]
  25. Johnson M, Bickel W. Within-subject comparison of real and hypothetical money rewards in delay discounting. J. Exp. Anal. Behav. 2002;77:129–146. doi: 10.1901/jeab.2002.77-129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kahneman D, Tversky A. Prospect theory: an analysis of decision under risk. Econometrica. 1979;47:263–291. [Google Scholar]
  27. Kim H, Shimojo S, O’Doherty JP. Is avoiding an aversive outcome rewarding? Neural substrates of avoidance learning in the human brain. PLoS Biol. 2006;4(8):e233. doi: 10.1371/journal.pbio.0040233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Kim H, Shimojo S, O’Doherty JP. Overlapping responses for the expectation of juice and money rewards in human ventromedial prefrontal cortex. Cereb. Cortex. 2011;21(4):769–776. doi: 10.1093/cercor/bhq145. [DOI] [PubMed] [Google Scholar]
  29. Knutson B, Adams CM, Fong GW, Hommer D. Anticipation of increasing monetary reward selectively recruits nucleus accumbens. J. Neurosci. 2001;21(16):RC159. doi: 10.1523/JNEUROSCI.21-16-j0002.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kosslyn SM. Aspects of a cognitive neuroscience of mental imagery. Science. 1988;240:1621–1626. doi: 10.1126/science.3289115. [DOI] [PubMed] [Google Scholar]
  31. Kosslyn SM, Shin LM, Thompson WL, McNally RJ, Rauch SL, Pitman RK, Alpert NM. Neural effects of visualizing and perceiving aversive stimuli: a PET investigation. Neuroreport. 1996;7:1569–1576. doi: 10.1097/00001756-199607080-00007. [DOI] [PubMed] [Google Scholar]
  32. Kringelbach ML, Rolls ET. The functional neuroanatomy of the human orbito-frontal cortex: evidence from neuroimaging and neuropsychology. Prog. Neurobiol. 2004;72(5):341–372. doi: 10.1016/j.pneurobio.2004.03.006. [DOI] [PubMed] [Google Scholar]
  33. Kuhnen CM, Knutson B. The neural basis of financial risk taking. Neuron. 2005;47(5):763–770. doi: 10.1016/j.neuron.2005.08.008. [DOI] [PubMed] [Google Scholar]
  34. Lagorio CH, Madden GJ. Delay discounting of real and hypothetical rewards III: steady-state assessments, forced-choice trials, and all real rewards. Behav. Process. 2005;69(2):173–187. doi: 10.1016/j.beproc.2005.02.003. [DOI] [PubMed] [Google Scholar]
  35. Lancaster J, Woldorff M, Parsons L, Liotti M, Freitas C, Rainey L, Kochunov P, Nickerson D, Mikiten S, Fox P. Automated Talairach atlas labels for functional brain mapping. Hum. Brain Mapp. 2000;10:120–131. doi: 10.1002/1097-0193(200007)10:3&#x0003c;120::AID-HBM30&#x0003e;3.0.CO;2-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Maldjian J, Laurienti P, Kraft R, Burdette J. An automated method for neuro-anatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. NeuroImage. 2003;19:1233–1239. doi: 10.1016/s1053-8119(03)00169-1. [DOI] [PubMed] [Google Scholar]
  37. McClure SM, Berns GS, Montague PR. Temporal prediction errors in a passive learning task activate human striatum. Neuron. 2003;38(2):339–346. doi: 10.1016/s0896-6273(03)00154-5. [DOI] [PubMed] [Google Scholar]
  38. McClure SM, Laibson DI, Loewenstein G, Cohen JD. Separate neural systems value immediate and delayed monetary rewards. Science. 2004;306:503–507. doi: 10.1126/science.1100907. [DOI] [PubMed] [Google Scholar]
  39. Nichols T, Brett M, Andersson J, Wager T, Poline J-B. Valid conjunction inference with the minimum statistic. NeuroImage. 2005;25:653–660. doi: 10.1016/j.neuroimage.2004.12.005. [DOI] [PubMed] [Google Scholar]
  40. O’Doherty J, Kringelbach M, Rolls E, Hornak J, Andrews C. Abstract reward and punishment representations in the human orbitofrontal cortex. Nat. Neurosci. 2001;4:95–102. doi: 10.1038/82959. [DOI] [PubMed] [Google Scholar]
  41. O’Doherty J, Buchanan T, Seymour B, Dolan R. Predictive neural coding of reward preference involves dissociable responses in human ventral midbrain and ventral striatum. Neuron. 2006;49:157–166. doi: 10.1016/j.neuron.2005.11.014. [DOI] [PubMed] [Google Scholar]
  42. Pagnoni G, Zink CF, Montague PR, Berns GS. Activity in human ventral striatum locked to errors of reward prediction. Nat. Neurosci. 2002;5(2):97–98. doi: 10.1038/nn802. [DOI] [PubMed] [Google Scholar]
  43. Penny WD, Holmes AP, Friston KJ. Random effects analysis, Human Brain Function. 2nd ed Academic Press; 2003. [Google Scholar]
  44. Peters J, Büchel C. Overlapping and distinct neural systems code for subjective value during intertemporal and risky decision making. J. Neurosci. 2009;29(50):15727–15734. doi: 10.1523/JNEUROSCI.3489-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Peters J, Büchel C. Neural representations of subjective reward value. Behav. Brain Res. 2010;213(2):135–141. doi: 10.1016/j.bbr.2010.04.031. [DOI] [PubMed] [Google Scholar]
  46. Pine A, Seymour B, Roiser JP, Bossaerts P, Friston KJ, Curran HV, Dolan RJ. Encoding of marginal utility across time in the human brain. J. Neurosci. 2009;29(30):9575–9581. doi: 10.1523/JNEUROSCI.1126-09.2009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Pylyshyn ZW. Mental imagery: in search of a theory. Behav. Brain Sci. 2002;25:157–182. doi: 10.1017/s0140525x02000043. [DOI] [PubMed] [Google Scholar]
  48. Quick SL, Shahan TA. Behavioral momentum of cocaine self-administration: effects of frequency of reinforcement on resistance to extinction. Behav. Pharmacol. 2009;20:337–345. doi: 10.1097/FBP.0b013e32832f01a8. [DOI] [PubMed] [Google Scholar]
  49. Schonberg T, O’Doherty JP, Joel D, Inzelberg R, Segev Y, Daw ND. Selective impairment of prediction error signaling in human dorsolateral but not ventral striatum in Parkinson’s disease patients: evidence from a model-based fMRI study. NeuroImage. 2010;49(1):772–781. doi: 10.1016/j.neuroimage.2009.08.011. [DOI] [PubMed] [Google Scholar]
  50. Schultz W. Predictive reward signal of dopamine neurons. J. Neurophysiol. 1998;80:1–27. doi: 10.1152/jn.1998.80.1.1. [DOI] [PubMed] [Google Scholar]
  51. Sescousse G, Redouté J, Dreher JC. The architecture of reward value coding in the human orbitofrontal cortex. J. Neurosci. 2010;30(39):13095–13104. doi: 10.1523/JNEUROSCI.3501-10.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Simmons W, Martin A, Barsalou L. Pictures of appetizing foods activate gusta-tory cortices for taste and reward. Cereb. Cortex. 2005;15:1602–1608. doi: 10.1093/cercor/bhi038. [DOI] [PubMed] [Google Scholar]
  53. Tobler PN, O’Doherty JP, Dolan RJ, Schultz W. Human neural learning depends on reward prediction errors in the blocking paradigm. J. Neurophysiol. 2006;95(1):301–310. doi: 10.1152/jn.00762.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Tobler P, Fletcher P, Bullmore E, Schultz W. Learning-related human brain activations reflecting individual finances. Neuron. 2007;54:167–175. doi: 10.1016/j.neuron.2007.03.004. [DOI] [PubMed] [Google Scholar]
  55. Wittmann M, Leland DS, Paulus MP. Time and decision making: differential contribution of the posterior insular cortex and the striatum during a delay discounting task. Exp. Brain Res. 2007;179(4):643–653. doi: 10.1007/s00221-006-0822-y. [DOI] [PubMed] [Google Scholar]
  56. Valentin VV, O’Doherty JP. Overlapping prediction errors in dorsal striatum during instrumental learning with juice and money reward in the human brain. J. Neurophysiol. 2009;102(6):3384–3391. doi: 10.1152/jn.91195.2008. [DOI] [PubMed] [Google Scholar]
  57. Zaghloul KA, Blanco JA, Weidemann CT, McGill K, Jaggi JL, Baltuch GH, Kahana MJ. Human substantia nigra neurons encode unexpected financial rewards. Science. 2009;323:1496–1499. doi: 10.1126/science.1167342. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES