Predicting Neurostimulation Responsiveness using fMRI

This is a supplementary dataset of the following paper:Yang, L. Z., Zhang, W., Wang, W., Yang, Z., Wang, H., Deng, Z. D., . . . Zhang, X. (2020). Neural and Psychological Predictors of Cognitive Enhancement and Impairment from Neurostimulation. Adv Sci, 7(4), 1902863. https://doi.org/10.1002/advs.201902863 fMRIParticipants were scanned on a 3.0-Tesla MR system (Discovery MR750, General Electric), with an 8-channel high-resolution radio-frequency head coil. Individual high-resolution T1 images were also acquired using a 3D BRAVO EDR FAST sequence (TR = 8.15, TE = 3.17, flip angle = 12°, Matrix = 256×256, slice thickness = 1mm). Functional BOLD images were acquired using the T2-weighted echo planar imaging sequence (36 axial slices, slice thickness = 4mm, TR = 2s, TE = 30ms, Matrix = 64×64, FOV = 220×220mm). Participants completed a task-based fMRI scanning using the false-belief task and a resting-state fMRI scanning. The false-belief task composed of two conditions: (1) reading stories describing false beliefs (BELIEF) and (2) reading stories describing false photographs and maps (PHOTO). The contrast between the BELIEF and PHOTO condition can identify brain regions involved in ToM and social cognition. The task script was modified from the script obtained from the Saxelab website (http://saxelab.mit.edu/superloc.php). See supporting information for the Chinese language material used. The resting-state fMRI comprised of 6-minutes functional imaging during which participants were required to close their eyes and had a rest. Behavior measure during online stimulationThe visual perspective-taking was measured using the ball-tossing task. In the task, the participant served as a virtual agent (the blue avatar) to play with the other two avatars (the red and the green avatar) on the screen. The participant’s visual perspective was either the same (the first-person perspective, 1PP) or different (the third-person perspective, 3PP) with the perspective of the blue avatar, the agent they represented (Figure 2A, 2B). There were one practice block and 4 test blocks — the practice block composed of 36 trials with all possible conditions. The subsequent 4 test blocks were in the balanced order with 36 trials in each block. In one-third of the trials, participants had to throw the ball—by a button press on a keyboard—to the red avatar by indicating whether the red avatar was on the left of the right side of the blue avatar (active condition). Then an animation of ball-tossing was followed by the response. In the other trials, participants had to indicate whether the avatar on his (or her) left or right side had the ball in hand (passive condition). If they made a correct answer within the 2 seconds, the avatar with the ball in hand threw the ball to the blue avatar, else, the ball was thrown to the other avatar. The practice performance was monitored to ensure that participants understood the task instructions. The view switching cost was defined as the relative IES difference between the third-person perspective condition and first-person perspective condition scaled by the baseline IES in the first-person perspective condition (Equation 1). The normalization could reduce outliers in individual analysis. The visual perspective-taking performance defined using response times and accuracy were also analyzed and presented in the supplementary material.