| ObjectiveSearching for the target from the environment involves multiple attention mechanisms. Maintaining task-set, overcoming the interference of distractor and shifting the focus need more subjective efforts, which involve the top-down mechanism. Besides, processing the salient stimuli needs less subjective effort involving more bottom-up mechanism. Theeuwes (1992) reported that when the salient distractor was present, searching more time was needed for a target. Furthermore, Mazaheri (2011) discovered that the more salient distractor corresponded to longer searching time. This indicated that the focus of attention was first occupied by the salient distractor and the top-down mechanism had difficulty to inhibit the attention capture. Schubo (2009) found the distractor presented alone rather than accompanied the target evoked a larger N1(140-170ms) and N2(220ms) components of ERP in the occipital region, and they were not influenced by the distractor salience, supporting the view of top-down mechanisms regulating the bottom-up process in the early phase. Ansorge (2011) found that when the clue contained characteristics of the target, the target could generate a larger N2pc (220-280ms) components in the occipital region indicated the participation of top-down mechanism. It can be speculated that the salient stimuli was first processed to capture attention by the bottom-up mechanism, followed by the top-down mechanism evacuating attention from the distractor. However, the interaction and the timing of these mechanisms are not clear enough and still need more study.This research was designed to explore the relations of the bottom-up and top-down mechanisms in visual search through graded the saliency of target.MethodsEighteen college students were recruited, but two of them were removed as their high EEG artifacts. The other sixteen college students (7males) aged between22and26years. All participants were right-handed and had normal corrected vision, having no color blindness, color weakness, nervous or mental illness. All subjects had not participated in similar experiments previously, and received Y40.00for their participation in consent.Each stimulus picture contained4color squares with edge length of1.09cm distributed evenly in four quadrants on a black background. The values of saturation and brightness of the color were240and120separately. The tones of the salient and the non-salient distractors were200(pink) and160(dark blue) separately with the presentation probability of0.50. The tones of low and high salient target were140(light blue) and135(sky blue); The presentation probability of salient and non-salient target were0.33. The view angle of each square was0.62°and the angle between the center of the most peripheral square and the fixation of the screen was2.78°.In this study, a two-way repeated-measure MANOVA was carried out for behavioral performance and ERP data:(salient distractor:YES, NO)*(target:none, low saliency, high saliency). Each session of the six experimental conditions had90pictures and each picture presented1500ms randomly in order, and the initial asynchronous for stimulus was2500ms. The subjects were asked to judge the four squares within the picture whether contained a target and pressed the button in left or right hand as soon as the target was found.EEG (bandwidth0.5-100Hz) was recorded using an ERP system developed in our lab, which could record behavior performance and EEG data at the same time. It was sampled from19electrodes mounted in an elastic cap according to international10-20system referenced to both earlobes. The impedance between scalp and electrode was smaller than10kΩ. Artifact above70μV and other artifact were eliminated manually. The ERP epochs were averaged off-line and included pre-stimulus activity of100ms and post-stimulus activity of1000ms.A two-way repeated-measure ANOVA was carried out for behavioral performance:(salient distractor: YES. NO)*(target: none, low saliency, high saliency). The reaction time and accuracy were analyzed using SPSS13.0with Greenhouse-Geissor correction. All ERP data were analyzed using statistical parametric mapping of F-values [SPM(F)] technique developed in our lab. The significant level was0.05.Results1Behavioral performance:The difference of the accuracy under different conditions did not reach the level of significance.For reaction time, the interaction between target and distracter was significant:F (1.99,29.79)=21.560, P=0.000. Either the salient distractor present or not, the simple effects of the target salience were significant:F(1.59,23.87)=102.494, P=0.000; F(1.20,18.03)=45.497, P=0.000. For the low salient target, significant difference of reaction time was found between the YES (552.57±82.32ms) and the NO (540.47±75.68ms) of distractor saliency:t(15)=-2.489, P=0.025. However, such difference of reaction time was not significant for the high salient target with the YES (527.14±77.84ms) and NO (527.77±78.31ms) of distractor saliency:t(15)=-0.162, P=0.873.2ERP and SPMIn the condition of2(distractor saliency: YES, NO)*3(target:none, low salient, high salient), the statistical parametric mapping of F-value of ERP revealed that:the significant interaction between target and distractor:the right frontotemporal area (700-750ms); the main effect of target: the occipital region and left prefrontal (350-450ms), and most brain regions exception of the bilateral prefrontal cortex (500-700ms); the distractor effect:the frontal-parietal areas and right temporal regions (350-450ms), left prefrontal (400-500ms) and the frontal-parietal areas (650-750ms). In the condition of2(distractor saliency:YES, NO)*3(target:low salient, high salient), the significant interaction between target and distractor:the left temporal region (450-600ms); the target effect:occipital region (350-450ms), right frontal-parietal and temporal island or/and left temporal region (500-700ms); the distractor effect:the right frontal-parietal temporal region (including the insular cortex)(350-450ms) and the left prefrontal (500-550ms).ConclusionBehavioral results from this study showed that the interference of salient distractor on searching for target would be weakened when the target became more salient, and then the interference may disappear when the target salience increases. So the salient target could get rid of interference of the distractor effectively. The result was in accordance with the conclusion of Geng (2010).The attention involves multiple functional subsystems (neural assemblies). The right ventral network involved reorienting caused by salient information, which is formed by the right inferior frontal gyrus, the former insular and the temporal-parietal cortex. The dorsal frontoparietal network involves selecting target and motoring response, which is formed by anterior cingulate cortex, frontal eye field and bilateral intraparietal sulcus. Based on the SPM of ERP that reveals the activation series of visual attention networks, we can divide the processing into three stages.(1) During the perception stage (350-450ms), the salient target only activates the visual area and the distractor impacts the right ventral salient network at the same time, with the two networks operating independently.(2) During the pattern classification stage (450-600ms), the left occipitotemporal region has a significant interaction effect suggesting it engaging the classification of targets and distractors as a familiarity processing network since the distractor and target both have a certain familiarity and competing each other.(3) During the executive stage (500-700ms), the salient target suppress and terminate the influence of the distractor in the left prefrontal (500-550ms) and its processing advantages keeps in the frontoparietal network and spreads to the right ventral attentional network (550-700ms). In addition, in the condition of the distractor present only, subjects need to press the button when the target does not exist, so the distractor also has task-relevance in this study. The distractor effect in frontal and parietal (650-750ms) only appeared in2*3analysis in a later stage, suggesting that it involves in monitoring response to the distractor. Significant target and distractor effects in the left prefrontal suggesting that it may play a role in recognition of stimulus of task-relevance.In summary, the rapid processing of the salient target not only benefits from the parallel processing of the target and distractor in the perception phase, but also from the effective classification by the familiarity network in the left temporal area in the stage of pattern recognition, which is an important stage for the target overcoming the interference from distractor. Results of this study suggest that attention process involving multiple neural assemblies and multiple streams of information processing. |
PDF Full Text Request