The Study On Global And Local Imagery Generation

A fundamental question of cognition and perception is how complex entities are processed, such as whether an entire visual scene is decomposed into or constructed by its parts. The global-local paradigm is a useful tool for assessing hemispheric asymmetry from neuropsychological and imaging data.The idea that imagery and perception share many of the same underlying mechanisms has been well established. As one of the most important cognitive abilities, mental imagery is the product of numerous distinct processes, and different combinations of these processes are used to accomplish different tasks, such as generation, maintenance, scanning and transformation, among which perhaps the most basic process is image generation. Kosslyn (1994) defined image generation as a complex and multi-componential process that consists of the temporary re-activation of representational units retrieved form long-term memory into a specialized processor, or "visual buffer".Despite a large number of imagery studies, only a limited number of research works have been carried out to investigate the generation of global and local imagery. Sui (2007) has investigated the generation of global-local imagery and its associated gender differences. The results showed that male adults generated global imagery faster than local imagery whereas no significant difference was found for female adults. To our knowledge, only one study has examined hemispheric asymmetry in global-local imagery. Farbrizio Doricchi (1994) concluded that multipart mental images are generated by the left hemisphere and skeletal images by both hemispheres. However, these studies were analyzed solely on the basis of behavioral measures (i.e. response time).The frequent use of event related brain potentials (ERP) recordings to examine global-local perception is supported by growing literatures, but with less attention in studies of mental imagery. Most of these studies asked the participants to generate visual mental images evoked in response to pictures, word expressions or sounds. The methodology in the present study derives from previous studies that provided clear and coherent results on the specific effects of different factors in image generation. In view of those issues, this dissertation implemented the following studies:1. Behavioral study of gender differences in global-local imagery generation. In this experiment,46 undergraduate students were selected as subjects, who acquired 20 animal pictures, according to which, experiments presented a question of global or local mental imagery,and then the generation time of global and local imagery was recorded by pressing buttons of the subjects. The results show that there were significant gender differences in the generation of global and local imagery generation, the global imagery generation was faster than the local one; the response of females were quicker than that of males, especially in the generation of local imagery generation. It might conclude that the global and local imagery generation results from the fulfillment of different processing, which shows significant difference because of sex.2. Global and local imagery generation and its gender difference:evidence from ERP. To investigate the neural correlates of global-local imagery, ERP were recorded while subjects were informed to generate the global/local imagery by presenting with the names (cue word for global or local target, such as"reading lamp"or "lamp shade") of objects (such as the electronic appliances and items form daily life).The results indicate (1) that the response time data was a delayed response to the generation of local imagery. In addition, ERP data mirrored the behavioral finding that P300 latency was faster for global targets than local targets during the 300-650ms range in both hemispheres. The amplitude was significantly greater in local imagery than in global imagery in the time window of 300-650ms with significantly increased activity over the left hemisphere for local stimuli, but no hemispheric difference for global targets was found. (2) ERP data indicated that the amplitude was significantly greater in local imagery in men than in women in the time window of 300-650ms with significantly increased activity over the left hemisphere for local stimuli for females.3. To analyze the organization of functional connectivity of event-related potentials (ERP) activity during the generation of global and local imagery, we evaluated the spatial patterns of ERP correlation between all possible pairs of electrodes (60) placed over the scalp of 28 healthy young normal subjects participating in a global/local imagery generation task, using two graph theoretical measures:the clustering coefficient and average path length. Results showed the functional network of global imagery generation got a shorter average path length, while that of local imagery generation got a larger clustering coefficient. Furthermore, the average distance between functional connected electrodes were longer in the generation of global imagery than local imagery. We suggested that the difference of functional connectivity patterns might reflect the different processing modes of the generation of global and local imagery that global imagery emphasized particularly on global integration, while local imagery on local specialization.4. The experiment demonstrates that global-local imagery generation can differ as a function of culture. Easterners (faster) and Westerners responded differently as they generated global image. The amplitude was marginally significantly greater in global imagery in Westerners than easterners in the time window of 300-650ms. It is suggested that imagery processes are influenced by culture. Western cultures have been found to organized objects by emphasizing rules and categories and to focus on salient objects, whereas people in East Asian cultures are more inclined to attend to the context. The cultural environment, both social and physical, shapes perceptual and imagery processes will be discussed.