Threat Detection in Nearby Space Mobilizes Human Ventral Premotor Cortex, Intraparietal Sulcus, and Amygdala

Abstract

In the monkey brain, the precentral gyrus and ventral intraparietal area are two interconnected brain regions that form a system for detecting and responding to events in nearby "peripersonal" space (PPS), with threat detection as one of its major functions. Behavioral studies point toward a similar defensive function of PPS in humans. Here, our aim was to find support for this hypothesis by investigating if homolog regions in the human brain respond more strongly to approaching threatening stimuli. During fMRI scanning, naturalistic social stimuli were presented in a 3D virtual environment. Our results showed that the ventral premotor cortex and intraparietal sulcus responded more strongly to threatening stimuli entering PPS. Moreover, we found evidence for the involvement of the amygdala and anterior insula in processing threats. We propose that the defensive function of PPS may be supported by a subcortical circuit that sends information about the relevance of the stimulus to the premotor cortex and intraparietal sulcus, where action preparation is facilitated when necessary.

Keywords: fMRI; peripersonal space; threat; virtual reality; visual looming.

Figure 1

Example of (A) an experimental trial and (B) an oddball trial.

Figure 2

Behavioral responses of VR threat experience questionnaire. The mean (n = 9) threat ranking scores of each stimulus and the standard errors are displayed (1 = low threat, 4 = high threat). DM = dark-skinned man, LM = light-skinned man, DO = dog, CH = child. Significant differences between ranking scores (p < 0.0083) are indicated with an asterisk *.

Figure 3

Network in response to the presence of approaching avatars. Results of the RFX ANOVA for avatar presence, FDR < 0.05. The outlines of the maximal probability maps of the ROIs are shown in white.

Figure 4

Additional visualizations related to the RFX ANOVA ROI analyses. For each ROI, the plot shows the mean beta value across threat conditions (DM, LM) subtracted with the mean beta value across threat control conditions (DMM, LMM) on the left (dark gray) and the mean beta value across no treat conditions (DO, CH) subtracted with the mean beta value across no-threat control conditions (DOM, CHM) on the right (light gray).

Figure 5

Network in response to presence of threatening vs. non-threatening avatars. Results of the RFX ANOVA threat perception contrast in yellow (initial threshold = 0.001, p (corrected) < 0.05), overlaid for display purposes on results with initial threshold = 0.005, p (corrected) < 0.05, in orange. The outlines of the maximal probability maps of the ROIs are shown in white.