hybrid models of motor control

models

• often used to explain how CNS performs functions

  • traditional models assign specific functions to distinct brain regions → failure to consider distributed nature of brain functions

advanced models of CNS function

• overcome constraints imposed by traditional feedforward and feedback models

  • hybrid models

hybrid models

advanced models that use distributed networks of brain regions to implement features of both feedforward and feedback models

multiple networks models

• use diff networks to plan, initiate, monitor, and regulated motor skills

  • cognitive, emotional and sensory networks

  • diff functs but converge and work together to decide how and if we will do motor movement

cognitive network

• uses network of cortex, basal ganglia, and thalamus to plan and initiate motor skills

emotional network

• network of limbic, basal ganglia, and brainstem areas to initiate actions in response to emotional stimuli

  • fear - fight or flight

sensory network

network of cerebellum, thalamus, and cortex to monitor and regulate motor skills

affordance competition hypothesis

• uses 2 distinct brain networks for motor planning

  • competition b/w diff motor skills to achieve task goal

• selection and specification network performed simultaneously (select b/w motor skills that are already being specified)

selection network

• decides what action (motor skill) to perform

  • includes temporal, parietal, and frontal cortex and cerebellum

specification network

• formulates how to execute actions (motor skills)

• includes occipital, parietal, and frontal cortex and cerebellum

internal models

• hybrid models that use efference copies of motor plans or commands to regulate motor skills

efference copies

• internal copies of outgoing motor plans or commands

• motor plans: specify body movements in ext coordinates (spatial)

• motor commands: specify temporal patterns of muscle activations

  • 2 variants are inverse internal and forward internal models

inverse internal models

• use efference copies of motor plans to regulate movements

• planner (P) in premotor cortex sends motor plans to controller (CT) in motor cortex

• controlled (CT) sends motor commands to controlled object (CO) in spinal cord to initiate movements of hands, arms, legs

• planner (P) sends efference copy of motor plans to inverse model (IM) in cerebellum)

• inverse model (IM) uses efference copies to predict sensory feedback

• inverse model (IM) compares predicted feedback with actual feedback (FB) to regulate motor skills

forward internal models

• use efference copies of motor commands to regulate movements

• planner (P) in premotor cortex sends motor plans to controller (CT) in motor cortex

• controlled (CT) sends motor commands to controlled object (CO) in spinal cord to initiate movements of hands, arms, legs

• controller (C) sends efference copy of motor commands to forward model (FM) in cerebellum

• forward model (FM) uses efference copies to predict sensory feedback

• forward model (FM) compares predicted feedback with actual feedback (FB) and sends result of comparison to controller (CT) to regulate motor skills

optimal feedback control

• proposes motor skills (motor behavior) result from interactions b/w brain and body

  • proposes optimal controller in motor cortex performs feedforward and feedback control

• task goals influence how feedforward and feedback control use sensory feedback

feedforward control

uses sensory feedback from body to transform motor plans to motor commands used initiate motor skills

feedback control

uses sensory feedback from body and environment to regulate ongoing movements

good consensus regarding several features of hybrid models

dynamic interactions, task goals, sensorimotor integration, neural networks

dynamic interactions

motor control involves interactions b/w individual (brain, body), environmental (perceptual), and task (cognitive) features

task goals

muscle activities and body movements ONLY modified by CNS if they improve outcomes specific to task goals

sensorimotor integration

cns combines multiple sources of sensory and motor info to optimize feedforward and feedback control

neural networks

distibuted networks of brain regions used for planning, initiating, monitoring, and regulating motor skills