tutorial i missed Flashcards
Correcting movement errors without sensory feedback
What effect does increasing the reward associated with a target have on eye movements towards that target?
It increases the velocity of the eye movement
Correcting movement errors without sensory feedback
What effect does blinking have on eye movements?
It pushes the eye, altering its trajectory
Correcting movement errors without sensory feedback
What problems do the above observations cause for the occulomotor system?
The motor commands that initiate eye movements are variable, due to variations in context and situation. The problem, then, is how the oculomotor system is able to fixate on a target accurately despite this variation in initial state.
Correcting movement errors without sensory feedback
What is corollary discharge and how does it deal with this problem?
Corollary discharge is an internal copy of an action command which predicts the sensory consequences of an action before sensory feedback is available. Corollary discharge is able to compensate for variability in the initial motor commands and produce accurate eye movements.
Correcting movement errors with delayed sensory feedback
What problem does delayed sensorimotor feedback cause for the motor system?
Slow sensory feedback can destabilize movements. Essentially, it results in continual over-adjustments, whereby the absence of feedback leads to additional movements, and then when the feedback arrives it shows the movement has gone too far, so a big counter-adjustment is made, and so on.
Correcting movement errors with delayed sensory feedback
How might a forward model solve this problem?
A forward model would generate a prediction of the position and velocity of a limb based on prior motor commands – this would enable feedback control to happen in real-time and avoid the instability associated with feedback delays.
Correcting movement errors with delayed sensory feedback
What additional advantage do forward models have?
Forward models can account for perturbations affecting not only the directly perturbed muscle, but also knock-on effects on other, connected muscles – in other words, forward models can take into account the more complex physical dynamics of the limb
Correcting movement errors with delayed sensory feedback
What evidence is there to suggest the cerebellum might generate such forward models?
TMS to the cerebellum disrupts the ability to make accurate movements to a target that has disappeared. Subjects made arm movements to the side and when a tone was played they had to make a forward movement to a remembered target location. If the movement was based just on the sensory feedback from the target then they should be inaccurate, as by the time they move their arm the feedback will be out of date. However, subjects are able to make accurate movements, which are disrupted by TMS to the cerebellum. Suggests the cerebellum computes the likely position of the arm in the future (i.e. when the movement is initiated) and compensates by adjusting the trajectory.
The effect of predicting sensory consequences of motor commands on perception
Why would the oculomotor system have a better idea of the position of the eye after an eye movement than before?
After making a movement, the brain is able to combine two sources of information – the prediction of the sensory consequences of an eye movement (provided by the forward model) and the sensory feedback after completion of the movement, providing a more precise estimate than if it had to rely on a single source of information.
The effect of predicting sensory consequences of motor commands on perception
What is the role of the parietal cortex in this process?
Neurons in the parietal cortex have been shown to remap their receptive fields before an eye movement occurs, based on a prediction of the sensory consequences of the movement – essentially, the parietal neurons provide a model of what the world will look like after the eye movement.
The effect of predicting sensory consequences of motor commands on perception
What would happen if we couldn’t predict the sensory consequences of making eye movements?
When we move our eyes across a scene the brain is normally able to predict that the stationary image will move across the retina and adjust our perception of the world accordingly. Without this ability the world would appear to move every time we move our eyes. This is equivalent to what happens when you move a camera while taking a picture – it blurs the image because the camera has no way to compensate for the motion
Combining predictions with sensory observations to produce a motor response
What is the optimal way to combine two sources of information?
Combine the two sources of information in a way that minimizes the variance of the resulting estimate. The idea is that the brain should incorporate a measure of uncertainty into its predictions.
Combining predictions with sensory observations to produce a motor response
How should each source of information be ‘weighted’ when being combined?
The information should be weighted in inverse proportion to the variance. So, the weighting will be higher for a source of information with lower variance, and the weighting will be lower for a source of information with higher variance.
Combining predictions with sensory observations to produce a motor response
How does the brain form a “belief” about the current state of the world?
By combining predictions about the sensory consequences of a movement with the actual measurements made at the time of movement.
Motor adaptation: Learning from sensory prediction errors
How might physical development of the body affect sensorimotor predictions?
It would change the relationship between motor commands and the actual motion of the limb, thus affecting the ability to predict the consequences of any particular motor command.
