Estimation of the timing of human visual perception from magnetoencephalography

Estimation of the timing of human visual perception from magnetoencephalography
复制标题

DOI:
10.1523/jneurosci.4343-05.2006
复制
发表时间:
2006-04-12
影响因子:
5.3
通讯作者:
Ohtani, Y
Ohtani, Y
中科院分区:
医学1区
文献类型:
--
作者:
Amano, K;Goda, N;Ohtani, Y

文献摘要

被引文献

相似文献

为了探索视觉感知的时间和潜在的神经动力学,我们分析了视觉刺激开始的手动反应时间(RT)与脑磁图(MEG)同时测量的诱发神经反应的时间过程之间的关系。视觉刺激是随机点从不连贯运动到连贯运动的转变,以及来自均匀场的色光栅的开始,这引起了不同皮质部位的神经反应。对于这两种刺激,从时间积分 MEG 响应跨越阈值(积分器模型)开始,可以准确预测中位 RT 随刺激强度(运动相干性或色彩对比度)变化的变化,并具有与刺激无关的检测后延迟。相比之下,从峰值 MEG 潜伏期或从非集成 MEG 响应跨越阈值(电平检测器模型)的时间开始,RT 的预测不太准确。积分器模型还可以至少部分地解释 RT 或对相同刺激的感知(命中/未命中)的试验间变化。尽管我们主要检查试验中平均数据的 MEG-RT 关系,但积分器模型甚至可以显示单次试验数据的一些相关性。当仅使用可能源自纹状皮层的早期视觉反应作为积分器模型的输入时,模型预测会恶化。我们的结果表明,对视觉刺激外观的感知是在纹外区域[MT(颞中视觉区域)周围的运动和V4(第四视觉区域)周围的颜色]建立的,类似于受试者对刺激做出手动反应之前的150-200毫秒。
To explore the timing and the underlying neural dynamics of visual perception, we analyzed the relationship between the manual reaction time (RT) to the onset of a visual stimulus and the time course of the evoked neural response simultaneously measured by magnetoencephalography (MEG). The visual stimuli were a transition from incoherent to coherent motion of random dots and an onset of a chromatic grating from a uniform field, which evoke neural responses in different cortical sites. For both stimuli, changes in median RT with changing stimulus strength (motion coherence or chromatic contrast) were accurately predicted, with a stimulus-independent postdetection delay, from the time that the temporally integrated MEG response crossed a threshold ( integrator model). In comparison, the prediction of RT was less accurate from the peak MEG latency, or from the time that the nonintegrated MEG response crossed a threshold (level detector model). The integrator model could also account for, at least partially, intertrial changes in RT or in perception (hit/miss) to identical stimuli. Although we examined MEG-RT relationships mainly for data averaged over trials, the integrator model could show some correlations even for single-trial data. The model predictions deteriorated when only early visual responses presumably originating from the striate cortex were used as the input to the integrator model. Our results suggest that the perceptions for visual stimulus appearances are established in extrastriate areas [around MT ( middle temporal visual area) for motion and around V4 ( fourth visual area) for color] similar to 150-200 ms before subjects manually react to the stimulus.