A distributed fMRI-based signature for the subjective experience of fear

Abstract

The specific neural systems underlying the subjective feeling of fear are debated in affective neuroscience. Here, we combine functional MRI with machine learning to identify and evaluate a sensitive and generalizable neural signature predictive of the momentary self-reported subjective fear experience across discovery (n = 67), validation (n = 20) and generalization (n = 31) cohorts. We systematically demonstrate that accurate fear prediction crucially requires distributed brain systems, with important contributions from cortical (e.g., prefrontal, midcingulate and insular cortices) and subcortical (e.g., thalamus, periaqueductal gray, basal forebrain and amygdala) regions. We further demonstrate that the neural representation of subjective fear is distinguishable from the representation of conditioned threat and general negative affect. Overall, our findings suggest that subjective fear, which exhibits distinct neural representation with some other aversive states, is encoded in distributed systems rather than isolated ‘fear centers’.

Angiotensin II regulates the neural expression of subjective fear in humans - precision pharmaco-neuroimaging approach

“…2E ). These findings were in line with our previous study (Zhou et al, 2021) showing that conditioned fear and subjective fear exhibit distinct neural representations.…”

“…The first 5 volumes of each run were discarded to allow for T1 equilibration. In line with our previous studies (Zhou et al, 2021) the remaining volumes were corrected for differences in the acquisition timing of each slice and spatially realigned to the first volume, and unwarped to correct for nonlinear distortions related to head motion or magnetic field inhomogeneity. The anatomical image was segmented into grey matter, white matter, cerebrospinal fluid, bone, fat, and air by registering tissue types to tissue probability maps.…”

“…However, despite the important role of the extinction paradigm to map fear processing in translational models and despite the important role of the amygdala in fear processing (e.g. Janak & Tye, 2015; Mihov et al, 2013) it has become increasingly clear that the subjective feeling of fear is not accessible in rodent models and neural indices restricted to isolated brain regions are insufficient to make inferences about mental processing in humans including the subjective experience of fear (LeDoux & Pine, 2016; Woo, Chang, Lindquist, & Wager, 2017; Zhou et al, 2021). In contrast, the subjective feeling of fear is the primary reason for patients to seek treatment and effects in this domain may represent the most relevant outcome to determine the potential of novel pharmacological treatments for fear (Taschereau-Dumouchel, Michel, Lau, Hofmann, & LeDoux, 2022).…”

“…This has the advantage of avoiding meta-reflective processes which may confound the actual mental process of interest and limit ecological validity. In the current study, we, therefore, employed a number of emotional processspecific multivariate signatures to decode the effects of LT on fear-related emotional experiences, including general negative affect (Chang, Gianaros, Manuck, Krishnan, & Wager, 2015), acquired threat (i.e., conditioned fear (Reddan, Wager, & Schiller, 2018)) and subjective fear experience (Zhou et al, 2021). The combination of these decoders allowed us to test whether and which specific fear-related processes are modulated by the AT1R system.…”

Putting the “mental” back in “mental disorders”: a perspective from research on fear and anxiety

“…Studying the connectivity between these different regions—as assessed by structural imaging based on diffusion, or resting-state fMRI data—may also predict individual differences in discordance and desynchrony [ 145 , 146 ]. Similarly, machine-learning algorithms trained to predict self-report, physiology and behavior [ 142 , 149 ] could also help us reveal the brain mechanisms associated with discordance and desynchrony. Studying such individual differences in brain processes might therefore help us better understand how discordance and desynchrony are associated with pathological conditions.…”