9:30 AM - 9:45 AM
[HDS10-03] Exploring brain regions associated with risk sensitivity and risk scrutiny attitudes toward tsunami evacuation behavior; an fMRI study using a simulation task
Keywords:disaster prevention, tsunami, evacuation behavior, decision making, fMRI
Understanding how humans make evacuation decisions is an important issue in facilitating evacuation from tsunamis in the event of earthquakes. We have developed tsunami evacuation simulation tasks (Figure 1: A,B) and found that risk sensitivity (the tendency to evacuate regardless of the risk of the scenario), risk scrutiny attitude (the tendency to evacuate under high-risk scenarios), and the tendency to perceive risk from sensory information rather than numerical information independently contribute to tsunami evacuation behavior.
This study attempted to elucidate the underlying neural basis by conducting fMRI experiments using the developed tsunami evacuation simulation task and attempting to measure brain activity during actual evacuation decision making from a tsunami. Risk sensitivity was positively correlated with emotional control traits in behavioral experiments, and emotional control traits were negatively correlated with brain responses during tsunami evacuation decision making in emotional response areas (anterior cingulate gyrus (ACC), insular cortex (INS)) (Takubo et al., in preparation). Therefore, we tested the hypothesis that risk sensitivity is related to emotional response areas (ACC, INS). In addition, previous study shows that the hippocampus (Hipp) was significantly activated in decision making based on numerical information (Palombo et al., 2019) and the sensory association cortex (TPJ) in decision making based on sensory information (Yomogida et al., 2014) for risk scrutiny attitudes. Therefore, we tested the hypothesis that Hipp would be relevant in numerical scenarios and TPJ would be relevant in sensory scenarios. Specifically, We added 20 control questions(Figure 1C) to the 40-trial simulation task we developed, for a total of 60 questions, and examined the correlation between each factor and the brain regions that increase in activity when making tsunami evacuation decisions.
METHODS: We conducted an fMRI experiment with 58 adults in their 20s (31 males and 27 females; 21.51 ± 1.30 (SD) years old). Participants completed a 60-trial tsunami evacuation decision-making task (20 numerical, 20 sensory, and 20 control scenarios (Figure 1C); 1Evacuate or 2Do not evacuate), and their responses and brain activity were measured using fMRI. Subjects who answered less than 60% correctly in the control scenarios were excluded from the analysis as dishonest respondents, and data from 55 subjects were included. SPM analysis was use, and multiple regression analysis was conducted for each of the sensory and numerical scenarios. The dependent variable was the estimated brain activity (blood flow change) per pixel. The independent variables were the average evacuation rate for each subject as the risk sensitivity, and the matrix product of the standardized average evacuation rate for each scenario multiplied by the subject's response data as the risk scrutiny attitude, for each sensory and numerical scenario.
RESULTS: In numeric scenarios, a significant regions positively correlated with risk scrutiny attitudes was found in the HIPP (p=0.012) (Figure 2). No significant regions were found to correlate with risk sensitivity. In sensory scenarios, a significant region negatively correlated with risk sensitivity was found in the dACC (p=0.008) (Figure 3). No significant regions were found to correlate with risk scrutiny attitudes.
DISCUSSION: The positive correlation between risk scrutiny attitudes and hippocampal activity in numerical scenarios is consistent with previous cognitive neuroscience research. The negative correlation between risk sensitivity and dACC in sensory scenarios is also consistent with our previous findings of a negative correlation between emotion regulation traits and risk sensitivity. This result indicates that low activity in the emotional response domain underlies risk sensitivity. The results also suggest that two different approaches to attitudes may be effective to foster disaster literacy.
This study attempted to elucidate the underlying neural basis by conducting fMRI experiments using the developed tsunami evacuation simulation task and attempting to measure brain activity during actual evacuation decision making from a tsunami. Risk sensitivity was positively correlated with emotional control traits in behavioral experiments, and emotional control traits were negatively correlated with brain responses during tsunami evacuation decision making in emotional response areas (anterior cingulate gyrus (ACC), insular cortex (INS)) (Takubo et al., in preparation). Therefore, we tested the hypothesis that risk sensitivity is related to emotional response areas (ACC, INS). In addition, previous study shows that the hippocampus (Hipp) was significantly activated in decision making based on numerical information (Palombo et al., 2019) and the sensory association cortex (TPJ) in decision making based on sensory information (Yomogida et al., 2014) for risk scrutiny attitudes. Therefore, we tested the hypothesis that Hipp would be relevant in numerical scenarios and TPJ would be relevant in sensory scenarios. Specifically, We added 20 control questions(Figure 1C) to the 40-trial simulation task we developed, for a total of 60 questions, and examined the correlation between each factor and the brain regions that increase in activity when making tsunami evacuation decisions.
METHODS: We conducted an fMRI experiment with 58 adults in their 20s (31 males and 27 females; 21.51 ± 1.30 (SD) years old). Participants completed a 60-trial tsunami evacuation decision-making task (20 numerical, 20 sensory, and 20 control scenarios (Figure 1C); 1Evacuate or 2Do not evacuate), and their responses and brain activity were measured using fMRI. Subjects who answered less than 60% correctly in the control scenarios were excluded from the analysis as dishonest respondents, and data from 55 subjects were included. SPM analysis was use, and multiple regression analysis was conducted for each of the sensory and numerical scenarios. The dependent variable was the estimated brain activity (blood flow change) per pixel. The independent variables were the average evacuation rate for each subject as the risk sensitivity, and the matrix product of the standardized average evacuation rate for each scenario multiplied by the subject's response data as the risk scrutiny attitude, for each sensory and numerical scenario.
RESULTS: In numeric scenarios, a significant regions positively correlated with risk scrutiny attitudes was found in the HIPP (p=0.012) (Figure 2). No significant regions were found to correlate with risk sensitivity. In sensory scenarios, a significant region negatively correlated with risk sensitivity was found in the dACC (p=0.008) (Figure 3). No significant regions were found to correlate with risk scrutiny attitudes.
DISCUSSION: The positive correlation between risk scrutiny attitudes and hippocampal activity in numerical scenarios is consistent with previous cognitive neuroscience research. The negative correlation between risk sensitivity and dACC in sensory scenarios is also consistent with our previous findings of a negative correlation between emotion regulation traits and risk sensitivity. This result indicates that low activity in the emotional response domain underlies risk sensitivity. The results also suggest that two different approaches to attitudes may be effective to foster disaster literacy.