Gain control of saccadic eye movements is probabilistic

Area of Science:

• Neuroscience
• Ophthalmology
• Cognitive Science

Background:

• Saccades are crucial for visual processing, but accuracy is limited by target uncertainty and motor noise.
• A known hypometric bias (saccades falling short) may stem from penalizing overshooting targets more than undershooting.

Purpose of the Study:

• To test the hypothesis that saccadic bias results from a cost function penalizing hypermetric errors.
• To investigate how target positional uncertainty influences saccadic accuracy and endpoint variability.

Main Methods:

• Systematically manipulated target positional uncertainty during saccade tasks.
• Analyzed saccadic endpoints, variability, and bias across different uncertainty levels.
• Examined the relationship between variability, bias, and the cost function asymmetry.

Main Results:

• Increased uncertainty led to greater endpoint spread and more hypometric errors.
• Saccades showed a systematic bias toward the average of target locations, indicating prior knowledge integration.
• Cost function asymmetry correlated with individual differences in correcting overshooting vs. undershooting errors.

Conclusions:

• The saccadic system employs a probabilistic-Bayesian control strategy to manage target uncertainty.
• This strategy statistically minimizes the expected cost of saccadic errors.
• Prior knowledge and asymmetric cost functions play key roles in saccadic planning and accuracy.