The neural component-process architecture of endogenously generated emotion

Abstract

Despite the ubiquity of endogenous emotions and their role in both resilience and pathology, the processes supporting their generation are largely unknown.We propose a neural component processmodel of endogenous generation of emotion (EGE) and test it in two functional magnetic resonance imaging (fMRI) experiments (N1/432/293) where participants generated and regulated positive and negative emotions based on internal representations, usin self-chosen generationmethods. EGE activated nodes of salience (SN), defaultmode (DMN) and frontoparietal control (FPCN) networks. Component processes implemented by these networks were established by investigating their functional associations, activation dynamics and integration. SN activation correlated with subjective affect, withmidbrain nodes exclusively distinguishing between positive and negative affect intensity, showing dynamics consistent generation of core affect. Dorsomedial DMN, together with ventral anterior insula, formed a pathway supportingmultiple generationmethods, with activation dynamics suggesting it is involved in the generation of elaborated experiential representations. SN and DMN both coupled to left frontal FPCN which in turn was associated with both subjective affect and representation formation, consistent with FPCN supporting the executive coordination of the generation process. These results provide a foundation for research into endogenous emotion in normal, pathological and optimal function.