Because subjects did not receive feedback during the experiments, their oculomotor performance could be assessed offline. First we removed the eye-blink artefacts from the eye position record and identified saccades by applying a velocity threshold criterion of
50°/s. see more During the fixation period the eye-signal was ‘nulled’ to compensate for drifts. A saccade appearing in a period of 0–1000 ms after the go signal (the disappearance of the fixation dot) was taken as an indicative saccade, whose direction reflected the subject’s decision. Within a trial a fixation break was defined when subjects made eye movements 1.5° or more away from the fixation point during a period of 3000 ms before the go signal. The analysis across the subjects revealed fixation breaks in 13.13% of the search and 17.21% of the control trials. The Wilcoxon test revealed no statistical difference in the fixation performance between both tasks, P = 0.1419. Furthermore, the Kruskal–Wallis test showed no significant difference in fixation breaks for the four search conditions, P = 0.7353. The average detection performance across all subjects and all conditions was 93.9% correct with a standard deviation of 1.6%. The detection accuracy for each event type was computed across all subjects; the mean accuracy and the standard error of the mean are shown in Fig. 1C. A one-way analysis of variance (anova) comparing
the different search conditions revealed no significant difference selleck screening library in the performance between conditions (F3,51 = 0.71, P = 0.5511). The behavioural task in this experiment allowed us to observe cortical BOLD activity associated with covert visual search to one of two retinal locations, which were kept constant independent of the eyes being directed
straight ahead, left or right relative to the head (Fig. 2A–D). To reveal the parieto-frontal network involved in the covert visual search independent of the different search conditions, we identified voxels in which the BOLD response evoked by covert visual search, GNE-0877 independent of the particular condition, exceeded BOLD activity found in the non-search control conditions. The four search conditions considered were obtained by having the eyes in three different positions (straight relative to the head, left and right) and the search array left or right of the fixation spot for gaze straight ahead (Fig. 2A–D). The group-level random-effect analysis revealed a bilateral search-related BOLD response in and around the IPS, in early and later visual cortical regions, unilateral activation in the right precentral region comprising the FEFs and the SEFs, as well as in the anterior insula bilaterally at a significance level of P < 0.001, 40 contiguous voxels. The reported clusters passed correction for multiple comparisons by applying a FDR criterion of 0.005 at the voxel level (Table 1).