Attentional cues differently affect eye and head latencies in head-unrestrained gaze shifts
AZ Khan1, G Blohm2, RM McPeek1, P Lefevre3
1Smith-Kettlewell Eye Research Institute,It is well established that eye movements and attention are closely linked which has led to the pre-motor theory of attention. Until now, this hypothesis was based on experiments studying eye movements in isolation. However, in natural conditions gaze saccades comprise an eye-in-head and the head-in-space component. Further it has been hypothesized that the head component of a gaze shift is double-controlled, i.e. through a gaze controller and a separate head controller. Therefore, we asked if attentional cues influenced head movements in combined eye-head gaze shifts and whether this was different than for eye movements.
Seven subjects made combined eye-head gaze shifts to targets either 30 deg to the left or to the right. Prior to the presentation of the target, a behaviorally irrelevant cue was flashed for 25 ms at the same or opposite location, at 50, 100, 200, 500 or 1000 ms stimulus onset asynchrony (SOA) times. We measured eye-in-head movements with a head-mounted Chronos video eye tracker at 400 Hz, head position and orientation with an active infra-red marker tracking Codamotion device, and neck muscle EMG (Neurolog) signals at 1000 Hz from the left and right sternocleido-mastoid and trapezius muscles. Saccade on- and offsets were detected using a 1000deg/s2 acceleration criterion and head movement onsets were detected using the EMG signals (first spike above 3*baseline SD) and, separately, using a 200deg/s2 acceleration threshold.
Comparing across congruent vs. incongruent cue-target locations, we found that the cue influenced the reaction times of both the eye and head components in a similar manner, with significantly longer reaction times for congruent compared to incongruent conditions at all SOAs. Interestingly, the head movement latency with respect to saccade onset was significantly shorter for congruent compared to incongruent cue-target locations.
The similar pattern of latency modulations across SOAs for saccades and head movements is in agreement with an attentional modulation of a combined head and eye gaze plan within the FEF-SC network. The fact that relative to the eye, the head begins moving earlier for congruent compared to incongruent cue-target locations suggests an additional role of attention on the head motor pathway. This is compatible with an early influence of attention on movement planning that differentially affects the saccade and head controllers.