Institute of Physics and Astronomy, University of Potsdam, Potsdam, D-14476, Germany.
Department of Psychology, University of Potsdam, Potsdam, D-14476, Germany.
Sci Rep. 2017 Oct 11;7(1):12958. doi: 10.1038/s41598-017-13489-8.
Fixational eye movements show scaling behaviour of the positional mean-squared displacement with a characteristic transition from persistence to antipersistence for increasing time-lag. These statistical patterns were found to be mainly shaped by microsaccades (fast, small-amplitude movements). However, our re-analysis of fixational eye-movement data provides evidence that the slow component (physiological drift) of the eyes exhibits scaling behaviour of the mean-squared displacement that varies across human participants. These results suggest that drift is a correlated movement that interacts with microsaccades. Moreover, on the long time scale, the mean-squared displacement of the drift shows oscillations, which is also present in the displacement auto-correlation function. This finding lends support to the presence of time-delayed feedback in the control of drift movements. Based on an earlier non-linear delayed feedback model of fixational eye movements, we propose and discuss different versions of a new model that combines a self-avoiding walk with time delay. As a result, we identify a model that reproduces oscillatory correlation functions, the transition from persistence to antipersistence, and microsaccades.
固视眼动表现出位置均方根位移的标度行为，随着时间滞后的增加，从持续到反持续的特征转变。这些统计模式主要由微扫视（快速、小幅度运动）形成。然而，我们对固视眼动数据的重新分析提供了证据，表明眼睛的慢成分（生理漂移）表现出平均平方位移的标度行为，这种行为在不同的人类参与者中有所不同。这些结果表明漂移是一种与微扫视相互作用的相关运动。此外，在长时间尺度上，漂移的均方根位移表现出振荡，这也存在于位移自相关函数中。这一发现支持了在漂移运动控制中存在时滞反馈的观点。基于早期的固视眼动非线性时滞反馈模型，我们提出并讨论了一个新模型的不同版本，该模型将自回避行走与时间延迟相结合。结果，我们确定了一个可以再现振荡相关函数、从持续到反持续的转变以及微扫视的模型。
