Fig 1.
Fixation dot appeared for 400ms, during which fixation was monitored. The stimuli (bisected line) were presented for 50ms. The interstimulus interval was 4000ms, during which the subject had to indicate which segment of the bisected line was longer by pressing left or right arrow keys or the down key for equal segments. Line1: mid-bisected (each segments 75 mm), Line2: right bisected (left segment: 75mm, right segment: 70mm), Line3: extreme right bisected (left segment: 80mm, right segment: 75mm), Line4: left bisected (left segment: 70mm, right segment: 75mm), Line5: extreme left bisected (left segment: 75mm, right segment: 80mm).
Table 1.
The scoring system of the attentional bias in the Landmark task.
The second column depicts the length of the left and right segment of the stimuli. The scores associated with the response in the various line length are shown in column 2–4.
Fig 2.
Reproducibility of the spatial bias.
The bias scores of the three consecutive measurements plotted for all twenty subjects. The subjects are ordered according to the spatial bias. Where there are only two data points, scores from two measurements were virtually inseparable.
Fig 3.
Correlation of the lateralisation of white matter microstructure with the behavioural spatial bias.
Statistical Images are overlaid on the FMRIB58_FA standard image. The mean FA skeleton, thresholded at 0.2, is depicted in green shades. Significant cluster is indicated in red-to-yellow (p<0.05, corrected for multiple correlations). A thickened version of the significant cluster is used to facilitate visualization. Positive correlation in the parietal lobe indicates leftward bias correlating with higher FA in the right hemisphere.
Fig 4.
The connectivity of the cluster showing correlation between FA and behavioural data.
In the upper section axial slices at standard space coordinates z = 46mm, 36mm, 26mm and 16mm are shown. In the lower section the cortical projections of the tracks are shown. FEF: frontal eye field, IPS: intraparietal sulcus. The binary cluster masks were used as seed mask for each subject. Five thousand streamline samples from each seed voxel were drawn to build up a connectivity distribution that was standardised by the total number of generated tracts, thresholded at 10% and binarized. Population connectivity maps were derived for controls by adding these masks together and thresholding at two (Pathways passing through the given voxel in at least two subjects).