Happiness matters: distributed brain patterns underlie different positive emotions in OFC

723 

Abstract Submission 

Alessandra Pizzuti¹, Sebastian Dresbach¹, Assunta Ciarlo², Michael Luehrs², Rainer Goebel¹

¹Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience, Maastricht, Netherland, ²Brain Innovation B.V., Maastricht, Limburg

Alessandra Pizzuti  
Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience
Maastricht, Netherland

Sebastian Dresbach  
Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience
Maastricht, Netherland

Assunta Ciarlo  
Brain Innovation B.V.
Maastricht, Limburg

Michael Luehrs  
Brain Innovation B.V.
Maastricht, Limburg

Rainer Goebel  
Department of Cognitive Neuroscience, Faculty of Psychology and Neuroscience
Maastricht, Netherland

Positive psychology is a promising branch in the field of psychology and the distinction of positive emotions receives increasing attention. Theoretical frameworks have proposed a "family tree" of 9 positive emotions that developed during evolution [1]. However, the neuroscientific basis of these emotions is poorly understood [2,3,4]. To shed light on this issue, a crucial aspect is to effectively elicit specific positive emotions. Unfortunately, using stock-photos has led to variable results [5]. Therefore, we utilized a novel, individualized approach to collect and rate positive emotions of memories. These memories were then used in fMRI experiments to investigate neural representations of positive emotions in healthy volunteers. Our insights may offer a promising path towards a biomarker of subjective wellbeing.

11 healthy volunteers participated in the experiment. Crucially, participants used the "Matter"-App [6] to collect images that could be associated with individual memories and rate to which degree each of 9 positive emotions (Fig1A) was present. Firstly, participants collected 3 peak-memories per emotion from their lives. Peak memories were used as reference for newly collected memories. Secondly, they recorded positive memories for 6 weeks to reach a minimum of 110 memories in total. From these memories, we selected a subset of 72 that best distributed the occurrence of each emotion. During the fMRI experiment, individual images from this set were presented for 5s while participants vividly recall the memory and the emotional content associated with it. After 5s, the image was replaced by a gray fixation dot with a yellow border while participants continue reliving the emotional memory for 15 seconds without the image cue (while keeping fixation on the dot). Individual trials were separated by 20s of rest periods indicated by the disappearance of the yellow border of the fixation dot. The setup ensured that the participants' images were not seen by the experimenters. Whole brain (f)MRI data (see Fig1B) was collected on a classical 7T Siemens Magnetom Plus scanner at Scannexus (Maastricht, NL) at 1.8 iso mm resolution and TR=1s. Data was analyzed as follows: slice time correction, motion correction [8], geometric distortion correction [9], high-pass filtering with 5 cycles [8]. FMRI data were co-registered to anatomical images using the boundary-based registration and registered to the MNI space in [8]. Finally, we computed a parametric general linear model (GLM) by using subjective ratings from the Matter App.

Our results from 4 participants show that viewing the images and "recalling" the associated positive memories elicited strong activity in both visual processing areas (VTC) and in emotion-related regions (Fig1C) such as orbital frontal cortex (OFC). Notably, the responses to only the recall phase show little visual activity, but still emotion related activations, as expected (Fig1D). These results indicate that participants engaged in the task and were able to recall the emotions. Fig2 shows the group-level responses to four individual emotions. Remaining emotions were not clearly differentiated. Averaging across subjects can indeed suppress emotion-specific patterns due to individual variation in distributed patterns. Our future efforts will involve the use of a multivariate pattern analysis that might be more effective in differentiating individual neural correlates of emotions.

With our novel approach which uses personal memories for eliciting positive emotions in a fMRI environment, we were able to show distinct responses to individual positive emotions in our cohort. This opens the door for in-depth investigations of positive emotions as the experimenter can rely on the participants information about what they felt in the scanner. Furthermore, this approach may aid patients to engage in positive memories during clinical neurofeedback studies using emotion regulation training for depression treatment [7].

Emotion and Motivation Other ¹

Imagery

Activation (eg. BOLD task-fMRI) ²

Other - 7T, positive emotions, high-field MRI, memory recall, parametric analysis