Fig. 1: Experimental paradigm and main analyses.
a Building a brain affective arousal signature. A whole-brain model (brain affective arousal signature, BAAS) was developed on the discovery cohort that underwent an arousal induction fMRI paradigm (study 1, n = 60) using the support vector regression algorithm and validated in study 2 (n = 36). b Evaluating the brain affective arousal signature. Based on the dimensional models of ‘core affect’, BAAS was evaluated in terms of generalizability in sixteen independent datasets (studies 3–18, n = 635) and in terms of specificity in nine independent datasets (study 2 and studies 19–26, n = 294). c Identifying the neurofunctional representation of affective arousal in the brain. Multivariate and univariate approaches were employed to determine the contribution of specific brain systems to predict subjective affective arousal. The thresholded affective arousal signature was compared with the conjunction of meta-analytic maps for positive and negative affect to further validate the biological plausibility of the identified affective arousal brain systems. Prediction analysis and conjunction analysis were next conducted to test performance of isolated brain systems in predicting subjective affective arousal experience. Finally, the neural representations between affective arousal and autonomic arousal signatures were evaluated to validate existed affective arousal-related domains. d Improving the specificity of neuroaffective signatures. Testing the specificity of two affective neural signatures (VNAS and VIDS) before and after BAAS response corrected. VNAS, visually negative affect signature from Čeko et al.34; VIDS, visually induced disgust signature from Gan et al.13. The icons were sourced from Pixabay under the Pixabay License or created in PowerPoint 2016. All icons are free to use in both commercial and noncommercial print and digital media.
