Two hundred and thirty-four healthy university students (144 females; average age = 19.87 years, standard deviation ^SD = 1.08) with no history of psychiatric or neurological disorders. The present study was a part of an ongoing research project (GBB project) which explored the associations among gene, brain, and behavior ^{26, 27}. This data set was analysis and some results were reported at our previous study ¹³. All participants had normal eyesight or corrected eyesight, no color blindness, and were right-handed (as indicated on the Edinburgh Handedness Inventory; ²⁸ ). Written informed consent was obtained from all the participants prior to the study. Both behavioral and MRI protocols were approved by the Institutional Review Board of Southwest University.

The face-word Stroop task was adopted as the experimental paradigm to measure emotional conflict for this study ^{12, 13}. In this task, participants need to see a face picture which was superimposed an emotion word, and they were instructed to identify the facial expression of the target face while ignoring the meaning of the words by pressing a button.

The target stimuli consisted of 5 male and 5 female face pictures, with either happy or sad expression, selected from the Chinese affective picture system ²⁹. Each face picture was superimposed two Chinese characters, “愉快” (which means “happy”) or “悲伤” (which means “sad”). The combinations of facial expressions and superimposed words yielded two conditions: a congruent condition (e.g., character meaning happy superimposed onto a happy face picture) and an incongruent condition (e.g., character meaning happy superimposed onto a sad face picture). The stimuli were programmed by E-Prime 2.0 software.

A total 120 trials consisting of an equal amount of congruent and incongruent trials were included in the formal experiment and another 24 trials was for practice. Stimuli were presented in pseudo-random order for avoiding repetition priming effect ³⁰. The timing and order of each trial was as follows: a fixation dot was presented for a specific duration (500 ms) followed by a blank screen of variable duration (300–500 ms). Then, the target face appeared for 1000 ms at the center of the screen. Participants had to respond within 1500 ms. The inter-trial interval (ITI) varied randomly between 800 ms and 1200 ms, with a mean of 1000 ms.

All behavior data analysis was implemented in SPSS 18. For each participant, we calculated mean accuracy and reaction time (RT) for each condition. The paired t test was performed on the accuracy and RT data, respectively. The difference between mean RT of incongruent trials and mean RT of congruent trials was defined as “emotional conflict effect” or “emotional interference effect” ^{12, 13}.

To adequate control for individual differences in emotional conflict control, all participants completed the Combined Raven's Test (CRT, Chinese version). The CRT was widely used for intelligence testing in China, and it has been proved to have good reliability and validity ³¹. The CRT comprises 72 nonverbal items and each item consists of a matrix with a missing piece that is to be filled in by selecting the best answer from 6 or 8 alternatives. The number of correct answers while completing the CRT test within 40 minutes was used as an index of general intelligence, and it was served as a covariate in our statistical analyses.

MRI scanning was conducted on a Siemens 3T scanner (MAGENTOM Trio, a Tim system) with an eight-channel phased array coil. For each participant, 242 functional images were acquired with a gradient echo type Echo Planar Imaging (EPI) sequence (echo time (TE) = 30 ms; repetition time (TR) = 2000 ms; flip angle = 90 degrees; slices = 32; slice thickness = 3.0 mm; slice gap = 1 mm; field of view (FOV) = 220 × 220 mm²; resolution matrix = 64 × 64; in-plane resolution = 3.4 × 3.4 mm² ; interslice skip = 0.99 mm). In addition, a high-resolution T1 weighted magnetization prepared gradient echo sequence (MPRAGE: TR/TE/TI = 1900/2.52/900ms, flip angle = 9 degrees, matrix = 256 × 256) anatomical scan also was acquired for registration purposes and anatomically localizing the functional regions. One hundred and seventy-six contiguous sagittal slices were obtained with 1 × 1 mm in-plane resolution and 1 mm slice thickness. Before the resting-state fMRI scanning, participants were instructed to keep their eyes closed without falling asleep, and to keep their head as still as possible during the scanning.

Image preprocessing was performed using statistical parametric mapping software (SPM8, http://www.fil.ion.ucl.ac.uk/spm) and using the data processing assistant for resting state (DPARSF, http://www.restfmri.net/forum/DPARSF). For each subject, the first 10 images were discarded due to instability of the initial MRI signal and adaptation of participants to the circumstance. The remaining 232 images were preprocessed, which included slice timing, head motion correction, spatial normalization, and smooth. Briefly, the 232 images were slice acquisition corrected, aligned to the middle images for head-motion correction. The various covariates including white matter, cerebrospinal fluid, and Friston 24-parameter were regressed out in order to reduce potential impact of physiological artifacts ³². Prior study has proved that regressing out Friston 24-parameter is more effective than other movement correction methods ³³. None of the subjects had more than 2.5 mm maximum displacement in x, y, or z translation and 2.5° of angular motion during the whole fMRI scan. Then, the corrected images were spatially normalized to the Montreal Neurological Institute (MNI) EPI template in SPM8 and resampled to 3×3×3 mm³voxels. The images then were spatially smoothed with a Gaussian kernel (full-width at half-maximum ^FWHM = 6 mm). In order to be consistent with previous, we selected 6 mm isotropic Gaussian kernel ¹³. Finally, the smoothed data was linearly detrended, and was filtered using a typical bandpass (0.01–0.08 Hz) to reduce the influences of high-frequency noise and low-frequency drift.

To explore whether the key region we identified in the ALFF-behavior analysis interacted with other brain regions to predict the emotional conflict, a seed-based whole-brain functional connectivity approach was conducted. We selected the AMY (MNI: 2, 14, 50) as the seed region according to previous study ¹³. The region which was associated with emotional conflict in ALFF was as seed region of interest. We created spherical ROI (radius = 8 mm) according to the peak of seed region identified in the ALFF-behavior analysis. For each participant, the correlation coefficients, between mean time series of all voxels with the seed region and the time series of each voxel of other brain regions, were calculated. The correlation coefficients were converted into Z scores using Fisher’s r-to-z transformation, which formed a z-functional connectivity map for each participant.

Then, we performed a partial correlation analysis to explore the functional connectivity that was related to emotional conflict. The gender, age, and IQ were treated as the confounding covariates, and the interference effect of emotional conflict was treaded as the covariate of the interest. The corrected cluster threshold was set at p < 0.05 (AlphaSim corrected for multiple comparisons, with a combined individual voxel p value < 0.01 with cluster size > 75 voxels).

The demographic information of the sample (age, gender, and Raven’s score) and detailed behavior data were reported in our previous study ¹³. The paired t test was performed on the accuracy and RT data to detect the emotional conflict effect, respectively. The emotional conflict effect was 21 + 2 ms (RT of congruent trials minus RT of incongruent trials). The emotional conflict effect was not influenced by gender and IQ.

To explore whether the AMY identified in prior ALFF-behavior analysis interacted with other brain regions to predict the emotional conflict, we performed a seed-based whole-brain functional connectivity analysis. The AMY (MNI: 2, 14, 50) was defined as the seed region of interest (ROI). We found that after controlling age, gender, and Raven’s scores, emotional conflict was positively correlated with functional connectivity strength between the AMY (the seed ROI) and right supplementary motor area (SMA) (MNI: 21, -33, 51; t = 3.70; cluster size = 130 voxels; p < 0.05 corrected) (Figure 1, Table 1).

Figure 1. Correlation between seed-based functional connectivity (FC) and emotional conflict. The FC between the AMY and the right supplementary motor area (SMA) was positively correlated with emotional conflict. Scatter plot between the AMY-SMA FC and emotional conflict is shown for illustration purposes.