Abstracts

Rationale: Recent lesion and neuroimaging studies have demonstrated the important role of the amygdale and the anteromedial temporal lobe in emotion recognition. In most of the studies, subjects were asked to select a verbal label that best described an emotion shown in each picture. This type of tasks might be insensitive to subtle deficits. Moreover, subjects were given unlimited time exposure to pictures in many studies. Therefore, we applied the rapid facial emotion recognition task which use two-alternative forced choice with time limited exposure to stimuli. In the current study, we analyzed the data using the signal detection theory (SDT) to calculate the sensitivity of anger detection in patients with temporal lobe epilepsy (TLE) who underwent the anteromedial temporal lobectomy (ATL). Methods: Twenty-one patients (left-ATL: 11, right-ATL: 10) and 20 healthy control subjects (CTRL) participated in this study. Six faces each of anger and happiness were morphed to create single faces expressing varying intensities of anger (strong: 75%, moderate: 50%, weak: 25%). These morphed faces were used as target stimuli. The procedure consisted of five blocks. In one block, the set of 96 stimuli were presented in random order. An individual face was presented for 67ms as a target stimulus and immediately masked by a neutral face, followed by a response selection screen. The participants were required to make a judgment on whether the target stimulus expressed anger or happiness by clicking either of the two buttons. Hits ("angry" responses to angry faces) and false alarms ("angry" responses to happy faces) were summed for each of the three angry faces. The sensitivity [d'=Z(hit)-Z(false alarm)] for anger detection was analyzed with SDT. This study was approved by the Ethics Committee of Tokyo Medical and Dental University. Results: A three (GROUP: right-ATL vs. left-ATL vs. CTRL) × three (STIMULUS: strong vs. moderate vs. weak) repeated-measures ANOVA examining the sensitivity of anger detection showed a significant main effect of STIMULUS (p < 0.01) and GROUP (p < 0.01) and interaction between STIMULUS and GROUP (p < 0.01). Post-hoc tests revealed that the sensitivity of anger detection in left-ATL group was significantly lower than in CTRL group for strong and moderate anger. Conclusions: The current study revealed that left-ATL group had difficulty in detecting anger in facial expressions. The left anteromedial temporal lobe could play an important role for anger detection in accordance with previous neuroimaging studies, which indicated the increased activation of the left amygdale when processing angry facial expressions. Applying SDT to the facial emotion recognition task was useful to clarify the subtle impairment which might be overlooked in previous lesion studies.