Anti-saccades

What it measures

The anti-saccade task measures voluntary control over a reflexive eye movement. A stimulus appears suddenly to one side and the reflex is to look at it. The task requires the patient to suppress that reflex and direct gaze to the mirror position opposite, where the information they need appears.

Two abilities are loaded at once: inhibition of the reflexive saccade, and generation of a voluntary saccade to a position holding no stimulus when the movement starts. Failures of the first show as gaze pulled to the distractor. Failures of the second show as movements that are late or land short.

Performance depends on frontal control of the saccadic system and on sustained attention. It improves through childhood and declines in later life.

How it is administered

The patient sits at a comfortable screen distance with habitual correction worn, head still. Room lighting stays moderate and steady, with no reflection on the panel.

Each trial starts with a central fixation mark. A dot then flashes briefly on one side. The target, a single character, appears briefly on the opposite side. The patient reports the character. A correct report means gaze reached the opposite side while the character was still on screen.

The clinician explains that the dot is the distractor and the character always appears opposite it, runs practice trials until the instruction is understood, then starts the recorded block and stops prompting, because prompting invalidates the latency values.

Stimulus durations and trial count are set before starting. Keep them the same across sessions, because a shorter character duration makes the task harder.

Reading the result

VisReed marks each trial correct or not and times it, then reports accuracy over the block and a mean response time over the correct trials.

Accuracy is the proportion of trials where the character was reported correctly. A low value indicates that attention followed the distractor instead of moving away from it.

The response time runs from the target appearing to the answer being given, so it includes the time the patient spends typing. It is a response time and not a saccadic latency. A longer time indicates slower voluntary orienting, or a more effortful strategy for suppressing the reflex.

Read the two together. A patient who is slow and accurate is trading speed for control. A fast patient with many errors is not inhibiting. Both differ from one who is slow and inaccurate.

Errors clustered at the start suggest the instruction was not understood. Errors increasing through the block suggest fatigue. Compare a patient with their own repeat testing, not with published anti-saccade figures, which come from eye tracking.

When it is used

The task is used when an oculomotor or attention difficulty is suspected in a child, alongside pursuit and convergence testing. It is used after acquired brain injury, and in conditions affecting frontal function, as one observation among many.

It also serves as a baseline before orthoptic training targeting saccadic control. A poor result contributes to a picture. It does not identify a condition on its own.

Limitations

This test is in beta. Treat the output as indicative and not as a settled clinical measure.

The response is a manual report, not a recording of eye position, so gaze direction is inferred. A patient who first looks at the distractor and then corrects can still report the character, and that trial counts as correct. Accuracy therefore overstates how well the reflex was inhibited.

The recorded time contains manual reaction time and is much longer than a saccade latency from an eye tracker. Screen refresh and input timing add further delay.

Viewing distance is not fixed, so stimulus eccentricity in degrees varies between sessions and between patients. Ambient light is not controlled.

Results depend on attention, understanding and motivation, so a poor score in a tired patient means little. The task does not replace laboratory oculography.