ERT TREAT Vision & Laboratoire de Psychologie et NeuroCognition, UMR 5105 CNRS-Université Pierre Mendès France, 1251 avenue Centrale, 38040, Grenoble, France.
Graefes Arch Clin Exp Ophthalmol. 2011 Feb;249(2):223-32. doi: 10.1007/s00417-010-1451-5. Epub 2010 Jul 23.
Bottom-up and top-down processes are involved in visual analysis of scenes. Here we examined the influence of top-down visual demand on natural scene processing.
We measured accuracy and response time in adults performing two stimuli-equivalent tasks. Unfiltered, low or high spatial frequency (SF) natural scenes were presented in central, left, or right visual fields (CVF, LVF, RVF). The tasks differed only by the instructed visual demand. In the detection task, participants had to decide whether a scene was present or not. In the categorization task, they had to decide whether the scene was a city or a forest.
Higher accuracy was seen for the LVF in the detection task, but for categorization, greater accuracy was seen for the RVF. The interaction between Task and SF revealed coarse-to-fine processing in the categorization task for both accuracy and reaction time, which nearly disappeared in the detection task. Considering the interaction of Task, VF and SF, a left-hemisphere specialisation (i.e., RVF advantage) was observed for the categorisation of HSF scenes for accuracy alone, whereas a LVF advantage was seen for all SFs in the detection task for both accuracy and reaction time.
Our results revealed that the nature of top-down visual demand is essential to understanding how visual analysis is achieved in each hemisphere. Moreover, this study examining the effects of visual demand, visual field presentation, and SF content of stimuli through the use of ecological stimuli provides a tool to enrich the clinical examination of visual and neurovisual patients.
自下而上和自上而下的过程都参与了场景的视觉分析。在这里,我们研究了自上而下的视觉需求对自然场景处理的影响。
我们在执行两个刺激等效任务的成年人中测量了准确性和反应时间。未过滤、低或高空间频率(SF)自然场景呈现在中央、左或右视野(CVF、LVF、RVF)中。任务仅通过指令的视觉需求来区分。在检测任务中,参与者必须判断场景是否存在。在分类任务中,他们必须判断场景是城市还是森林。
在检测任务中,LVF 的准确性更高,但在分类任务中,RVF 的准确性更高。任务与 SF 之间的交互显示,在分类任务中,准确性和反应时间都存在从粗到细的处理,而在检测任务中,这种处理几乎消失。考虑到任务、VF 和 SF 的交互,仅在分类任务中,HSF 场景的分类存在左半球专业化(即 RVF 优势),而在检测任务中,所有 SF 的准确性和反应时间都存在 LVF 优势。
我们的结果表明,自上而下的视觉需求的性质对于理解每个半球中视觉分析是如何实现的至关重要。此外,通过使用生态刺激物来检查视觉需求、视野呈现和刺激物 SF 内容的影响,本研究为丰富视觉和神经视觉患者的临床检查提供了一种工具。