Abstract:
In recent years, most research on phenotype prediction from functional connectivity (FC) of the brain has been based on resting state (RS) data, where participants lie in the scanner without any particular task or any external stimulation. However, one of the major drawbacks of this approach is that, in the absence of a task, RS FC is influenced by the spontaneous thoughts of the participant and is therefore badly standardized. To address this limitation, naturalistic viewing (NV), during which participants are presented with movie clips while inside the scanner, has been suggested as an emerging tool for the study of individual differences. By closer mimicking conditions under which the brain naturally operates, NV promises to capture more ecologically valid neuronal responses and thereby increase prediction accuracies. Furthermore, by choosing stimuli to match the phenotype of interest, it might be possible to manipulate brain state to further increase predictive power. I will discuss recent results from our group addressing the question whether phenotype prediction based on NV is indeed superior to RS based predictions.
