Directing gaze at blind parts of the visual field

Past Event Date: 


Manfred Mackeben

Meeting room: 

Room 204 - Main Conference Room

Event Category: 

Event Type: 


Behavior in the visually oriented world often goes along with scanning eye movements that have a tight relationship with attention. Movements of the eyes generally follow movements of attention which can be evoked by a visual stimulus (or cue).

In a clinical setting, patients with restricted visual fields can learn to extend their field of view by making eye movements into the damaged part of their visual field. Examples are homonymous hemianopia and visual field restrictions, as in advanced glaucoma or retinitis pigmentosa. I've been involved in two studies showing that such patients can indeed learn to make eye movements that extend their functional field of view.

The question is: 

How is this possible? How can attention be attracted by a stimulus they cannot see?

There are two attentional mechanisms: Transient (TA, exogenous) and sustained (SA, endogenous) attention. TA requires a stimulus in the outside world, it is quick and reflex-like, while SA generated in the brain, is slow and under control of volition. Since TA depends on an external visual stimulus, it is not a candidate for guiding EMs into a blind part of the visual field. However, SA is not dependent on visual input, as it can be directed to an empty part of the visual field by an effort of will. This was already described by von Helmholtz in the late 19th century.

SA has been researched extensively in humans and in primates and there are a variety of brain regions that evoke attentional orienting. I will discuss findings from experiments on the superior colliculus, the posterior parietal cortex, and the frontal eye fields.

It is important to note that such consciously controlled EMs bring the intact fovea to the target location, which makes previously hidden objects visible. After the next EM, these objects may disappear again from the visual field but may persist in a remembered image of the visual surround that works with another time constant than primary vision. This way, goal-directed EMs could be made to targets that do not depend entirely on primary visual input.  This hypothetical mechanism could support  enlarging the field of view by learning.