跳視是一種共軛式之眼球運動,為最快速之眼球運動,其功能為快速將眼球中央凹對準物體以得到清晰的視覺。非共軛凝視則為兩眼往相反的方向轉動之眼球運動方式,其目的為將目標物影像對準左、右眼對應之網膜位置。主程序關係廣泛被用於跳視及非共軛凝視之動態特性分析,跳視區間和振幅成正比而峰值速度則為振幅之指數函數並於30-40度時呈現飽和,非共軛凝視之峰值速度則遠小於跳視。傅立葉分析亦被使用於跳視之量化研究,和時域不同的是頻域分析之特性和整個軌跡相關。先前的研究曾使用逆向方法利用位置軌跡之傅立葉轉換函數及系統模型之脈衝響應函數重建跳視之輸入控制訊號。在本文中,我們應用獨立成分分析(ICA)技術來分解跳視及非共軛凝視之成分。結果顯示脈衝成分在步級成分之前啟動,此結果和以前的研究結果相似。另外,非共軛凝視之調適成分亦可經由ICA之分析結果呈現。結論為ICA可以有效地應用於跳視及非共軛凝視之控制成分分解,並提供另一種眼球運動之研究方式。
Saccade is believed to be the fastest eye movement, which is conjunctive and functions to redirect the fovea of the retina to the object of interest. Vergence, on the other hand, are disjunctive eye movements that two eyes move in different directions, which in turn bring the images of the target for two eyes into their corresponding retinal positions. Main sequence relations have been used for describing the dynamics of the saccadic and vergent eye movements. Saccadic duration is approximately proportional to its amplitude, whereas the peak velocity and amplitude relation could be described as an exponential function with saturation occurred at 30-40°. Main sequence relations are also used for quantitative analysis of vergence eye movements that the peak velocity is significantly smaller than saccades. Fourier analysis has also been used in quantitative investigation of saccades. Unlike the temporal analysis for the main sequence relations, spectral analysis relies on the whole trajectories of saccades. An inverse method was also used by previous investigators to reconstruct the input signal of a saccade from the fourier transform of the padded and mirrored position profile and the impulse response of the system model. In this paper, we present the application of independent component analysis (ica) to decompose the components of saccade and vergence eye movements. The results show that the pulse component is activated before the step for both vergence and saccade, which is consistent with previous findings. Also the effect of accommodative activation can be observed for vergence using ica. In conclusion, ica is useful in decomposing the control components for the saccade and vergence from ensembles of recorded profiles, which provides an alternative method for eye movement investigation.