Malcolm George L, Silson Edward H, Henry Jennifer R, Baker Chris I
School of Psychology, University of East Anglia, Norwich, United Kingdom.
Laboratory of Brain and Cognition, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, United States.
Front Hum Neurosci. 2018 May 8;12:189. doi: 10.3389/fnhum.2018.00189. eCollection 2018.
We can understand viewed scenes and extract task-relevant information within a few hundred milliseconds. This process is generally supported by three cortical regions that show selectivity for scene images: parahippocampal place area (PPA), medial place area (MPA) and occipital place area (OPA). Prior studies have focused on the visual information each region is responsive to, usually within the context of recognition or navigation. Here, we move beyond these tasks to investigate gaze allocation during scene viewing. Eye movements rely on a scene's visual representation to direct saccades, and thus foveal vision. In particular, we focus on the contribution of OPA, which is: (i) located in occipito-parietal cortex, likely feeding information into parts of the dorsal pathway critical for eye movements; and (ii) contains strong retinotopic representations of the contralateral visual field. Participants viewed scene images for 1034 ms while their eye movements were recorded. On half of the trials, a 500 ms train of five transcranial magnetic stimulation (TMS) pulses was applied to the participant's cortex, starting at scene onset. TMS was applied to the right hemisphere over either OPA or the occipital face area (OFA), which also exhibits a contralateral visual field bias but shows selectivity for face stimuli. Participants generally made an overall left-to-right, top-to-bottom pattern of eye movements across all conditions. When TMS was applied to OPA, there was an increased saccade latency for eye movements toward the contralateral relative to the ipsilateral visual field after the final TMS pulse (400 ms). Additionally, TMS to the OPA biased fixation positions away from the contralateral side of the scene compared to the control condition, while the OFA group showed no such effect. There was no effect on horizontal saccade amplitudes. These combined results suggest that OPA might serve to represent local scene information that can then be utilized by visuomotor control networks to guide gaze allocation in natural scenes.
我们能够在几百毫秒内理解所看到的场景并提取与任务相关的信息。这一过程通常由三个对场景图像具有选择性的皮质区域支持:海马旁回位置区(PPA)、内侧位置区(MPA)和枕叶位置区(OPA)。先前的研究主要关注每个区域对视觉信息的反应,通常是在识别或导航的背景下。在这里,我们超越这些任务,研究场景观看过程中的注视分配。眼动依赖于场景的视觉表征来引导扫视,进而引导中央凹视觉。特别是,我们关注OPA的作用,它:(i)位于枕顶叶皮质,可能将信息输入到对眼动至关重要的背侧通路的部分区域;(ii)包含对侧视野的强视网膜拓扑表征。参与者观看场景图像1034毫秒,同时记录他们的眼动。在一半的试验中,从场景开始时起向参与者的皮质施加一列500毫秒的五个经颅磁刺激(TMS)脉冲。TMS应用于右半球的OPA或枕叶面部区(OFA),OFA也表现出对侧视野偏向,但对面部刺激具有选择性。在所有条件下,参与者通常做出从左到右、从上到下的整体眼动模式。当TMS应用于OPA时,相对于同侧视野,在最后一个TMS脉冲(400毫秒)后,向对侧视野的眼动扫视潜伏期增加。此外,与对照条件相比,对OPA施加TMS使注视位置偏向远离场景的对侧,而OFA组没有这种效应。对水平扫视幅度没有影响。这些综合结果表明,OPA可能用于表征局部场景信息,然后视觉运动控制网络可以利用这些信息来指导自然场景中的注视分配。