Neural Network Models of Saccade Generation and Spatial Accuracy

A computational modeling study by Peter Ford Dominey of the University of Southern California presents neural network models for oculomotor control. The models explain spatiotemporal transformations for reflexive saccades, voluntary control via cortical and basal ganglia functions, and learning mechanisms for saccade sequences. The work includes an analysis linking model parameters to saccade dysfunction in Huntington's disease patients.

Use Cases

• Modeling spatiotemporal transformations for saccade generation based on described neural network architectures.
• Simulating voluntary control of eye movements based on the described cortico-striatal and basal ganglia models.
• Analyzing mechanisms for maintaining spatial accuracy based on the introduced dual-mechanism theory.
• Studying sequence learning in motor systems based on the model combining sensory input with efferent copies.
• Investigating parameter sensitivity in neural models to understand dysfunction, as demonstrated for Huntington's disease.

Strengths

• Models are based on known anatomy, providing a biologically grounded computational framework.
• The description introduces specific theoretical mechanisms, including a dual-mechanism theory for spatial accuracy.
• Analysis connects model parameters to dissociable clinical dysfunction in two patient groups.

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

University of Southern California

Collection Method

Theoretical/computational modeling study.