Cognitive visual field rehabilitation through strategies of coordinated gaze motor skills

Zangemeister WH, Poppensieker K, Hoekendorf H, Shaker Verlag, Aachen 1999

Original text language: German

This paper links  re-view with pre-view of our work,-   going back to  gaze-saccadic strategies in healthy humans and hemianopic patients, connecting them with the complex viewing machinery necessary for vision and “reading”  of static, and dynamic pictures.  The visual system, particularly gaze movements  guide our behavior in natural environments. Many animals including humans have evolved a non-homogeneous retina to optimize information transmission and serially sample visual scenes by saccadic eye movements. Such gaze movements introduce high-speed retinal motion and decouple external and internal reference frames. Recently, it has been demonstrated  that perceptual decisions can be made in world- i.e. gaze, rather than retinal- coordinates.

Gaze saccades, very fast movements, are more often than not synkinetic coordinated fast eye and –frequently-  head movements. This means their internal latencies within the  central nervous system must be  flexible,  due to: i. anticipation-prediction of the subject;  ii. randomness of  internal and environmental stimuli; iii. dynamic differences and limitations of the fast eye compared to the slower head movements.

Anticipation-prediction plays a major role for  saccadic gaze movements. It permits us to find one  or more targets of interest; we can  connect multiple targets within a region of interest, a ROI,  through concatenation of sequences of fixations to “view” and understand a picture.  These sequences are called scanpaths (Noton & Stark 1971).

Their characteristics in space and time are defined by the subject´s top-down predictive knowledge of the environment, or a particular scene/ picture. Bottom up mechanisms are interwoven in this process such that a picture may -to some extent-  “steer” the sequence of fixations.

In normal life, pictures are most often dynamic scenes rather than static two- or three-dimensinal pictures. Gaze fixations occurring during dynamic visual input are therefore performed through smooth pursuit gaze movements and gaze saccades. Under natural conditions, visual sensitivity during gaze saccades is reduced. Gaze coordination of eye and head, extending between total prediction and pure randomness is the most important factor when executing scanpaths to input pictures and dynamic scenes. Fixational micro-eye-movements, in particular microsaccades,  contribute to perception and intimate understanding of  fixation of static pictures .