Disconnection of ipsilateral FEF from the attention network in strabismic amblyopia during coherent motion task

Abstract

Purpose : Anecdotally it has been suggested that the amblopic eye of strabismic amblyopes is less able and slower to drive eye movments than the fellow eye. Thus we compared speed threshold performance and brain activation during motion salience and voluntary saccade tasks in adult strabismic amblyopes and controls.

Methods : Psychophysical speed thresholds for target detection were measured in 8 strabismic amblyopes and 8 normal participants prior to fMRI scanning of a motion salience task and a task requiring voluntary saccades to horizontal stimulus position changes. Activation and functional networks between ROIs including parietal-frontal attention network, Frontal Eye Fields and V1 were examined.

Results : Significantly impaired speed thresholds for target detection were seen through the strabismic amblyopic eye compared to thresholds through the fellow eye or control eyes when determining the direction of motion of the coherently moving dots task. BOLD activation in V1, IPS and FEF was also weaker for both tasks, following viewing through the amblyopic eye compared to activation through the fellow or control eyes. Correlational connections between key nodes of the visual attention network activated by the amblyopic eye were abnormal in the motion salience task, with the FEF ipsilateral to the amblyopic eye isolated from all other nodes. By contrast, the functional connections of this network were normal for the voluntary saccade task.

Conclusions : Lower motion thresholds and less BOLD activation of the parietofrontal network through the amblyopic eye indicate a requirement for longer exposure time for discrimination of the moving target during amblyopic eye driven movements compared to those through the fellow eye or control eyes. A specific deficit in functional connectivity was also seen between V1 and the frontal eye field ipsilateral to the strabismic eye during the motion salience task but not during the voluntary saccade task.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.