final Flashcards
forward model
simulates behavior of the body in response to motor commands and captures causal relationship b/t action and consequence
-“predictor”
inverse model
simulates behavior of the motor apparatus and estimates motor commands required to achieve a desired state in advance of the movement and based on current state and corollary discharge
“controller”
describe internal model as part of a mechanism for internal feedback loops
- forward dynamic model provides a constant closed loop feedback system for online adjustment
- together with the inverse model
predictive and anticipatory control in terms of forward model
- predictor (forward model) uses the efference copy to predict the sensory feedback, which allows for anticipatory control of things like grip control
- eg ketchup
- prediction happens due to sensory delay which would be reactive control
sensory confirmation and cancellation in terms of forward model
- sensory signals from the environment during movement are compared to the predicted signals that the forward model predictor predicts from the efference copy
- if they match, the signal is interpreted as reafferent
- if they do not match, it is interpreted as exafferent information
- eg why external stimulation is more intense than self-produce
- but anything that disrupts this comparison (rotation of robot, delay of stimulus) can create a mismatch again in signals
- eg self produced force matching is inaccurate b/c you cancel out some of the feedback and have to create more force to feel it on your finger
internal forward model and action agency
- ability to distinguish an action as self-produced based on comparison of afference to efference copy
- eg schizophrenia patients have symptoms where they are not the agent of their own control
- could be due to a. generating inadequate internal predictions or b. impaired comparison of predicted and actual
reafference principle/internal model at the single neuron level
recorded from vestibular afferents in monkeys
- no difference b/t active and passive movements
- vestibular signal not attenuated on the single neuron level during active movement
- when recording from vestibular nuclei, there was a attenuation of the signal for head movements
vestibular signals during head movements
-active: vestibular signal is attenuated
passive head movement: codes for movement
-simultaneous passive and active movements: proprioception and vestibular afferent signal compared to reafference in vestibular nuclei
-active movement if vestibular attenuation, passive if no attenuation
how is an internal model updated for learning?
- brain learns to expect unexpected signals
- monkeys introduced to unexpected signal (causes error of attribution) over learning trials supressed the vestibular signal more and more
- internal model is reclaibrated
- after effect shown when unexpected signal removed
explain how the CNS determines whether a muscle is actively engaged in balance
it does if the efference copy of the motor command is congruent w/ the sensory signals
- is a balance response when congruent
a. active balancing vs. b. passive balancing where robot does for you - subject doesn’t know they aren’t actually balancing, so this is unconsciously controlled
- when it’s passive movement, the signals are incongruent w/motor response (because of robot) so there’s no balance response
division 1 of visual streams
1st: what vs. where
what=inferior temporal, where=post. parietal
-distinction is based on stimulus features
-both based on conscious perception
-how the stimulus is processed is important
division 2 of visual streams
- what vs. how
- ventral (inferior temporal)=what, dorsal (post. parietal)=how
- visual processing for online control of action
- separate systems to serve visual perception and action
eg of same info different visual system
- use info about shape and size to know it’s a remote
- also to know how to pick it up and how to shape the hand to grasp it
- what vs. how
blindsight
“seeing” what you can’t see
- evidence in monkeys w/visual cortex lesion
- they could still avoid objects and grasp objects and moving objects
- ie this is mediated through subcortical pathways
dorsal vs. ventral streams
both streams go through the visual cortex, ventral goes to inferior occipitotemporal cortex and dorsal to post. parietal
- ventral responsible for object recognition (textures, orientation, etc.)-conscious
- dorsal responsible for spatial recognition (depth perception, location, on-line actions)
evidence for what/how division
patient DF-visual agnosia w/ventral stream damage
- unable to recognize size and orientation of objects
- but preserved dorsal stream
posting: can’t accurately orient card for slot but can put it in the mail slot - can’t estimate size of objects by matching hand size but can go out and grab the object
Patient RV
- damage to dorsal stream (PPC)
- interrupted visuomotor processing for visually guided actions
- optic ataxia-can’t reach for obects accurately
- can recognize objects, can describe orientation of slot
- cannot draw well b/c uses visual info , can’t put card in slot
- grip aperture perception accurate, not accurate when picking up object
how are illusions used in visual systems studying
- illusion fools perception but not grip aperture
- allows for distinguishing b/t memory based and real time control of actions
how are patients w/optic ataxia still able to accurately perform grasping movements?
evidence for real-time vs. memory based control: patients w/optic ataxia (dorsal stream damage) who are still able to perform accurate grasping movements based on memory and internal predictive model
hand eye coordination during movement
- eye movement then initiation of reaching
- proprioceptive and visual information used to calculate appropriate motor output and efference copy
- error computed against efference copy to correct movement