Brain networks under uncertainty: A coordinate-based meta-analysis of brain imaging studies

Abstract

In recent years, uncertainty has been extensively studied as a core factor in anxiety models. However, it remains unclear whether there is a stable brain circuitry to cope with uncertainty. Addressing this yet open question, we first distinguish uncertainty into three different states: risky, ambiguity, and threat anticipation. Then, we performed three meta-analyses of fMRI studies to identify those regions that are commonly activated by the three domains using activation likelihood estimation (ALE). The overlapping analyses of the three ALE maps revealed major conjunctions of the risk decision making, ambiguity decision making, and the threat anticipation in specifically the right insula. Contrast analysis further confirmed this finding. In addition, different uncertainty states also have different brain networks involved. Specifically, a large number of brain regions in the frontal-parietal cortex were recruited under ambiguity state, while subcortical gray matter regions were recruited under risk decision making, and the bilateral insula were closely associated with threat anticipation. Additionally, we assessed the co-activation pattern of identified regions using meta-analytic connectivity modeling (MACM) to investigate the potential network underlying the relationship of three domains. The MACM analysis further confirmed that different uncertain states have specific brain network basis. We concluded that the right insula serves as a convergent brain region for brain regions recruited for different uncertain states, and its co-activation pattern also corresponds to the brain network of the three uncertain states. This study is a preliminary attempt to further uncover the brain circuitry of anxiety models with uncertainty at their core.