Wang et al. 10.1073/pnas.0709664104. |
Fig. 3. Schematic illustration of five stimuli used in experiment 6.
SI Methods
Subjects.
The subjects in all experiments were 18 to 32 years old, and all had normal or corrected-to-normal vision. In the fMRI study, informed consent issued by the local ethics committee was obtained from the participants.Stimuli and Methods.
In all behavioral experiments but experiment 7, the stimuli were drawn white on a black 21-inch computer screen with an average stimulus luminance of 6.87 cd/m2 and a background luminance of .03 cd/m2 at the viewing distance of 90 cm. In all of the experiments, subjects were instructed, with emphasis, to maintain their eyes on the fixation point throughout a trial, and to respond as accurately and quickly as possible. No feedback was given.In experiments 1 and 8a, subjects detected whether there was a quadrant different from the rest by pressing a button on a two-button pad. In experiments 2-4 and 6, subjects performed a discrimination task to judge which quadrant was different from the rest by pressing the correspondence button on a four-button pad. In experiments 5 and 8b, subjects in one group detected inside or outside, and the other group detected near or far, by pressing a button on a two-button pad.
In experiment 7, a Gerbrand four-field T-scope was used to present the pairs of stimulus figures drawn in black ink on white paper cards; two figures in each stimulus display were arranged vertically and located at either the left or the right side of the fixation point; subjects looked at the fixation point on a preexposure field and pressed a button to receive a 5-ms presentation of a field containing one of the three stimulus pairs, followed by the immediate reappearance of the preexposure field; and the intensity of illumination of the target field was adjusted for each subject to keep an overall probability of reporting "different" at ~50%. The sequences of trials (e.g., the location of an odd quadrant in experiments 1-4, 6, and 8a; the sequences of inside and outside, and near and far in experiments 5 and 8b; and the disk's location of being upper or lower in experiment 7) were randomized and counterbalanced across the subjects when needed.
Samples of photographic reductions of the five stimuli used in experiment 6 are shown in SI Fig. 3. SI Fig. 3d was modified from SI Fig. 3a in the way that, in its odd quadrant, two of the four right angles were rotated 180° to join the other two to form two squares; otherwise, they remain the same as those in the other quadrants. SI Fig. 3e was designed so that the odd quadrant contains a larger hollow square formed by four right angles, but the rest each contain four right angles randomly distributed. Thus, both SI Fig. 3 d and e represent the same topological discrimination based on holes.
fMRI Study
The subjects were scanned by a Siemens 3T Trio scanner with a standard single-channel head coil. In the fMRI study, the stimuli were presented through an LCD projector onto a rear projection screen, which was located behind the participant's head inside the magnet bore. The stimuli were presented centrally at the viewing distance of about 60 cm through an LCD projector onto a rear projection screen, which was located behind the participant's head inside the magnet bore. A T2*-weighted gradient-echo echo planar imaging (EPI) sequence was used, with a matrix size of 64 ´64, a field of view (FOV) of 200 mm ´ 200 mm, a slice thickness of 4 mm (1 mm interslice gap), and a repetition time/echo time/flip angle = 1,500 ms/29 ms/90°. Twenty-five axial slices were acquired to cover the whole brain. A blocked design was used in the functional scan, in which the blocks of activation or baseline stimulus were counterbalanced in order, and each block lasted 32 s. High-resolution anatomical images were collected by using a T1- weighted, three-dimensional gradient-echo sequence (3D MPRAGE; 1 mm ´1.3 mm ´ 1 mm resolution) for registration of functional images and localization of brain activation. Standard preprocessing, including realignment, spatial normalization and smoothing with Gaussian spatial kernels of 6 mm, and voxel-based statistical analysis were performed by using SPM2 (Wellcome Department of Cognitive Neurology, London). To allow population inference, two-stage analyses were performed for voxel-based statistical analysis. On the first level, linear contrast of interest (topological vs. local properties) was calculated to generate a contrast parameter estimate map for each subject. Afterward, the ensuing contrast images of the first level were submitted to the second level for random effects group analysis by using a one-sample t test, to obtain a statistical parametric map of the t statistics for each voxel. The average time course was extracted from significant peaks of activation clustered around the ROI identified in the group analysis in individual subject scans, within standard deviations (e.g., about 15 mm in condition A and about 10 mm in condition B). The percent signal changes of the activation stimulus relative to the baseline stimulus was calculated after removal of its linear trend, and then it was analyzed with a one-sample t test.