Sensorimotor Learning in Response to Errors in Task Performance.

Area of Science:

• Neuroscience
• Motor Control
• Cognitive Psychology

Background:

• The human sensorimotor system learns from prediction errors and performance errors.
• Mechanisms of performance error learning are unclear, often confounded with prediction errors.

Purpose of the Study:

• To investigate motor adjustments driven solely by performance errors, isolated from prediction errors.
• To differentiate the strategies underlying motor adaptation to performance errors of varying magnitudes.

Main Methods:

• Induced performance errors by shifting reach targets while matching intended and actual motion kinematics.
• Analyzed motor adjustments in response to isolated performance errors across three experiments.
• Investigated the influence of error size on adaptive strategies.

Main Results:

• Motor adjustments to performance errors relied on deliberative, volitional strategies, not implicit learning.
• Large performance errors engaged goal-directed, model-based control (prefrontal cortex-striatum pathways).
• Smaller performance errors utilized model-free stimulus-response learning (motor cortex-striatum pathways).

Conclusions:

• Performance errors drive strategic motor adjustments, dissociating based on error size.
• Distinct neural pathways support different error-driven learning strategies.
• Faster re-learning suggests retrieval of previously learned strategic compensation, independent of limb-error history.