Every day we take in large amounts of information from the external world, and we also synthesize representations of things or situations that we have not perceived through our senses. The ability to distinguish between a memory that contains representations from external world and a memory representing an imagined picture is necessary to make sense of the surroundings. This process is called reality monitoring. In the present study we aimed to confirm the existence of the reality monitoring network as reported by previous studies. Further, we wanted to extend these findings by investigating the effect of stimuli aversiveness on the reality monitoring processes and its neural correlates. 

Twenty-five subjects were included in the study after passing a somatic and psychiatric health screening. The subjects first completed an encoding task of 80 trials outside the scanner. Small descriptions of either an object or a situation (two or three word sentences) were presented on a computer screen. Immediately after the description was shown, a frame that was either empty or containing a picture related to the description was shown for three seconds. The subjects were instructed to look at the picture in the frame or imagine a relevant picture when the frame was empty. The subjects were then instructed to consider whether the pictures were “Unpleasant” or “Not unpleasant” by choosing between the two alternatives on the computer screen. A retrieval task was carried out as Blood-Oxygen Level Dependent (BOLD) fMRI data was collected. During this task the participants were presented with small descriptions that were either presented during the encoding task or they were new. The subjects were to decide whether they previously had viewed a picture associated with the description (a V trial), whether they had imagined a picture associated with the description (an I trial) or whether the description was entirely new (an N trial). The subjects completed a total of 140 randomly presented trials during two runs (20 trials of each category and 20 baseline trials). T2*-weighted functional MRI images were collected on a 3T General Electrics Signa HDx scanner. Data were analysed using SPM8.

Overall, most of the trials were considered neutral, and this was true within both the I and the V conditions. Fewer I trials than V trials were considered aversive. The response times were longer in I compared to the V for the aversive trials, and there was a trend for the same effect for the neutral trials. There were no significant differences in response time between neutral and aversive trial. The analysis of the retrieval task behavioural data revealed a higher accuracy rate for aversive trials in the I than the V, while there was no effect for neutral trials. An ANOVA for the corresponding response times showed a main effect of source of encoding where responses were shorter in V than I trials. In paired tests this difference was significant for neutral trials. Paired tests of emotional content within source showed a difference between aversive and neutral trials for I. Successful retrieval and discrimination between sources of encoding generated activations in the left posterior precuneus. Activations of the anterior cingulate were also present. An effect of stimuli aversiveness on brain activation was present in mediolateral prefrontal cortex and the precuneus, indicating a stronger effort of these regions during retrieval of source memory linked to aversive stimuli.

In summary, activation patterns in reality monitoring networks were replicated from earlier studies. Further, the results suggest that activations in overlapping networks are increased for aversive stimuli compared to neutral stimuli.