Adaptive control of movement deceleration during saccades

Area of Science:

• Neuroscience
• Computational Neuroscience
• Ophthalmology

Background:

• Accurate eye movements (saccades) are crucial for reading and visual perception.
• The brain learns from visual errors to improve saccade accuracy.
• A dual-process model proposes fast (low retention) and slow (high retention) learning mechanisms for adaptation.

Purpose of the Study:

• To identify neural signatures of hypothesized fast and slow saccade adaptation processes.
• To investigate how visual error adaptation is distributed across saccade phases (acceleration and deceleration).
• To characterize the retention properties of adaptation in different saccade phases.

Main Methods:

• Experimental manipulation of visual errors during saccade tasks.
• Analysis of motor commands during saccade acceleration and deceleration phases.
• Application of state-space models to characterize adaptive processes.

Main Results:

• Single visual errors induced adaptation primarily in the saccade deceleration period.
• Repeated errors led to adaptation in both acceleration and deceleration periods.
• Deceleration adaptation was rapid but poorly retained, while acceleration adaptation was slower but more robust to forgetting.
• State-space models supported a shared adaptive state plus two distinct processes (fast/deceleration, slow/acceleration).

Conclusions:

• Saccade adaptation involves distinct fast and slow learning processes affecting different phases of the eye movement.
• These processes exhibit differential sensitivity to errors and varying retention capabilities.
• The findings support a two-state model of motor learning in the context of visual-guided saccades.