The Visual System

Question 1

What did Thorpe et al (1996) show?

Answer 1

Using ‘go/no-go’ categorisation, participants were asked whether a briefly presented image had an animal in it or not.

This was done whilst measuring ERPs using implanted intracerebral electrodes

ERPs differed significantly after 150ms in correct go vs no-go responses, showing the necessary visual processing to distinguish between animal and distractor was <150ms

Response was stronger in frontal electrodes suggesting it’s role in inhibiting inappropriate repossess with no-go decisions

Question 2

Why is photo transduction referred to as an ‘amplification cascade’? (rods)

Answer 2

The absorption of a single photon (from a wide bandwidth of light) is sufficient to cause a significant change in membrane conductance in rods.

The steps of amplification:

Step 1: activated rhodopsin molecule to activate approx 100 transducins.

Step 2: the activated tranducins activates PDE which in turn catalyse the breakdown of up to 6 cGMPs into GMP.

Question 3

How do the ‘on’ and ‘off’ pathways differ between rods and cones?

Answer 3

Rods:
Do not have separate on and off

The rod bipolar communicate with off centre retinal ganglion cells via amacrine cells.

For the OFF RGC, the amacrine cells act directly, whereas to stimulate the ON RGCs, the amacrine cells do it indirectly by synapsing onto a cone ON bipolar cell.

Cones:
Have 2 separate pathways - 1 for on and 1 for off.

Cone bipolar cells make direct synapses with RGCs.

Question 4

Which cells mediate lateral inhibition?

What are the similarities and differences of lateral inhibition for an ON centre and OFF centre RGCs?

Answer 4

a) Horizontal cells
b) Similarities:

Horizontal cells are excited by increased glutamate release from cones, and inhibited by an inhibitory mechanism believed to involve GABA.

Responsible for the surround antagonism in the RGC receptive field

Differences:

With on centre RGCs horizontal cells act directly on the cone

With off centre RGCs the horizontal cells act at the synapse

Question 5

What are the functions of lateral inhibition?

Answer 5

a) contrast enhancement, increased dynamic range, reduces redundancy in the signal, facilitates constancy (by looking at border contrast and not absolute brightness)

Question 6

What are the inputs to the LGN and what are they for?

Answer 6

Approx. 10% come from RGCs

Brain stem inputs (30%) - mainly cholinergic and control the flow of information to the cortex.

Local inhibitory cells (30%) - GABAergic intrageniculate interneurons which receive direct retinal input and project to the LGN relay cells (feedforward inhibition) to strengthen inhibition and ensure precise and spatially organised Res that resemble those of RGCs.

Cortical (30%) - corticofugal inputs involved in enhancing the inhibitory surround of the LGN to best reflect its interpretation.

Question 7

Describe the key organisational features of the primary visual cortex.

Answer 7

Retinoic mapping

Ocular dominate patterns:
Inputs from the 2 eyes are segregated in layer 4 in alternating sections. Neurons here are monocular
In layers 2 and 3, the signals from he 2 eyes are combined, although neuron activation shows eye preference.

Orientation columns and hyper columns:
Simple cells with same orientation preference are grouped into orientation columns
A set of orientation columns covering all orientations = a hyper column

Question 8

What has cytochrome oxidase staining in V1 and 2 revealed about anatomical segregation?

Answer 8

Showed evidence of some anatomical segregation

V2 - Some blobs in the thin bands were shown to be colour selective (and project to V4)

V2 - Some in the thick bands were shown to be motion selective and projected to area MT

Cells in the pale bands receive input from inter blobs in V1 and are connected to V4 - thought to be involved in processing form.

Question 9

How did rod opsins evolve?

What determines an opsins molecular properties?

What’s the difference in sensitivity between cone and rod opsins?

Answer 9

a) Rods evolved after cones! - following duplications and divergence of the original SWS opsin gene. Photoreceptors expressing the Rh1 opsin gene eve loved to become rods.
b) The primary amino acid sequence - peak spectral sensitivity can be shifted. It was discovered that the amino acids at sites 122 and 189 are the main contributors to the functional differences between rods and cones.
c) There is no difference. The perceived higher sensitivity of rods is thought to be due to prolonged time in the activated state –> more time for G protein activation –> greater amplification

Question 10

What is the difference between single and double opponent cells?

Answer 10

Single (DeValois, 1966)
Cells with opposed inputs based on wavelength alone - centre input is different to surround input. = L+/M+ (luminance), L-/M+ (red/green), S-/L+M+ (blue/yellow) - koniocellular layers

Double (Conway and Livingstone 2006)
Only found in V1 and beyond. These are the basis for colour constancy. Elongated RFs with side by side spatially antagonistic regions.

Question 11

What are the opponent channels?

Who proposed the idea of opponent processing?

Answer 11

a)
Red- Green channel = L-/M+ = midget (P) ganglion cells
Blue yellow channel = S-/L+M = bistratified ganglion cells
Luminance = L+M = Magnocellular ganglion cells

b) Ewald Herring 1982

Question 12

What is chromaticity?

What are it’s 2 independent qualities?

Answer 12

a) Quality of colour independent of brightness
b) Hue - colour appearance is determined by the dominant wavelength

Saturation - determined by excitation purity. Depends on the mount of white light mixed with the hue. Pure hues are fully saturated - no white light is mixed in.

Question 13

What’s the difference between protanopia and protanomaly?

Answer 13

Protanopia = Lack of LW cones = can’t distinguish between certain colours.

Protanomaly = shifted LW peak.

Question 14

How do we perceive depth?

Answer 14

By binocular disparity - refers to the difference in image location of an object seen by the right and left eyes.

Some cells detect disparities in front of the plane of fixation = Near cells

Others respond to disparities beyond the plane of fixation = Far cells

Question 15

Properties of neutrons in the IT cortex?

Answer 15

Large receptive fields sometimes extending across the midline, usually including the fovea.

Clusters of face cells have been identified along the ventral temporal cortex - shows it has a role in processing more complex stimuli

RFs are not retinotopically organised

Magnitude of response is almost always greater to stimuli presented at the fovea - reflects its importance in object recognition.

Retinal size and precise location in the RF doesnt affect response

Responses can be modified by attention (e.g as shown with binocular rivalry experiments) and training etc…

Question 16

What evidence is there for population coding of orientation?

What factors increase accuracy of population coding?

Answer 16

a) Adapt to a grating and then measure the threshold contrast required to detect a vertical test grating. Repeat this for different orientations.

When plotted as a function of adapting grating orientation, declining threshold with increasing orientation implies population coding.

b) no of units. 
The average response strength 
orientation bandwidth
response variability 
(Vogels 1990)

Question 17

What are the differences between Baum’s and Meyer’s loops?

Answer 17

Baum’s: Travels back through the parietal lobe and gives info about the inferior part of the visual field.

Meyer’s: Travels back through the temporal lobe. Gives info from the superior part of the visual field.

think these are just part of the optic radiations - temporal and parietal

Question 18

What are the mechanisms of adaptation to a decrease in illumination in rods?

Answer 18

Adaptation allows a large range of illuminations to be perceived.

The two main features of adaptation are:

1. Decreased sensitivity
2. Accelerated response recovery

Ca2+ dependent mechanism…

1. Ca2+ binds to guanylate cyclase activating protein, preventing it from binding to GC, thus slowing it’s functioning. - slow the opening of gates
2. Ca2+ binds to calmodulin, allowing it to bind to cGMP gated channels, redoing their affinity for cGMP, causing some gates to close. - prevents the opening of gates.
3. Only occurs with longer adaptation - to prolong the response and increase sensitivity. Ca2+ binds to recoverin, allowing it to bind to rhodopsin kinase and decrease its activity.

Ca2+ independent mechanism:
Involves the light dependent increase in steady state PDE activity (Burns and Arshavosky 2005)

Question 19

What are the mechanisms of phototransduction termination?

Answer 19

2 step process to inactivate rhodopsin:

1. phosphorylation by rhodopsin kinase
2. Binding of arrestin to this phosphorylated form

Dissociation of all trans retinal and conversion to retinol by retinal dehydrogenase

GTP-ase mediated inactivation of transducin (the g protein) and PDE (the effector protein), accelerated by GTPase activating protein GAP - speeds up hydrolysis

Decrease in Ca2+ following hyper polarisation activates the CaBP, guanylyl cyclase activating protein (GCAP), which binds to and increases the activity of GC - faster cGMP synthesis - increased cGMP

Question 20

Why can cones generate photopigment quicker than rods?

Answer 20

Cones have their own RPE independent mechanism of regenerating 11-cis retinal - Muller cells supply cones with 11-cis retinol, which have the necessary enzymes to convert to 11-cis retinal (rods don’t)

N.b. the attachment of a new 11-cis retinal is the rate limiting step in resetting the system.
The mechanism in the RPE involves a series of enzymatic reactions.

Question 21

What are cardinal cells?
Who proposed their existence?
What type of coding do they do?

Answer 21

a) Cells that respond univariantly to the same external pattern e.g. faces - they are neurons selective for specific elements. AKA grandmother cells.
b) Barlow (1972) - The Neurone doctrine - he went on to say that we only have cardinal cells for important features.
c) sparse coding - a population of cells responding less as the selectivity increases.

Question 22

Advantages and disadvantages of colour vision?

Answer 22

Advantages:
Foraging for fruit and young leaves in mature foliage
Signalling sexual readiness
Improving recognition of a natural scene

Disadvantages:
Can hinder camouflage detection - Morgen et al 1992 found that dichromats could detect colour camouflaged objects that trichromats couldn’t.

Question 23

What have Britten et al added to our understanding of Area MT functioning? - Are neurones in area MT actually responsible for the global perception of motion?

Answer 23

Britten, Newton and Movshon 1989

They trained monkeys to report the direction of motion, either in the neurones preferred direction or null direction
Proportion of dots moving coherently
Found very little difference between neuronal and psychophysical

Salzman, Britten and Newsome 1990
Performed a similar experiment, but this time stimulated neurones in Area MT
Microstimulation shifted the psychophysical function to the length, meaning that the monkey was more likely to report that the dots were moving in the preferred direction.

Question 24

What proposition did Ungerlieder and Mishkin make in 1982?

Explain.

Answer 24

Ventral VS Dorsal streams for visual processing.

Ventral - Temporal - Object identification - affected by extra retinal factors - parvocellular - Produces visual agnosia when lesioned

Dorsal - Parietal - Object location in visual space and visually guided motor behaviour - Magnocellular - produces optic ataxia when lesioned

Question 25

Describe the ocular dominance columns of PVC (primary visual cortex)?

Answer 25

Input from the two eyes is separated into separate columns within layer 4 of the PVC.

Layers 2 and 3;
At these levels the inputs from each eye are combined, producing the basis for stereopsis (sensation of depth perception)

Question 26

What evidence is there for the presence of cardinal cells in the IT cortex?

Answer 26

Desimone et al 1984 - recorded from single neurones in IT cortex of monkeys, whilst presenting different stimuli and found cells that selectively responded to;
Faces, Hands, broader object properties e.g. shape, colour, texture etc…

Afraz et al 2006 - stimulation of area IT to bias face perception

Tsao et al 2006 - microelectrodes in area showing 97% of visually responsive cells were selective for an faces.

Schalk et al 2017 - Japanese man with epileptic activity over the FFA - was seeing faces in everyday objects - stimulation of FFA is sufficient to cause this! Implanted brain electrodes for his epilepsy could be stimulated.

Question 27

Where have imaging studies shown there to be a face selective region of the cortex?

Answer 27

Right mid fusiform area - aka fusiform face area (FFA) - identified by Kanwischer et al 1997 using fMRI.

Question 28

What is the tilt after effect?

What happens if the bar viewed after adaptation is beyond that of the tuning curve?

Answer 28

The tilt of an object is skewed in the opposite direction after a particular orientation is adapted to. This is due to decreased sensitivity of neurones which have been exposed to a particular orientation - if there is overlap in tuning curves between the adapted neuron and another orientation then an error occurs in estimation.

no illusion

Question 29

Whats the difference between V1 ad MT orientation selective cells?

Answer 29

V1 neurones are sensitive to motion that is perpendicular to their orientation
The aperture problem - if no edges or textures are present, the direction and speed of motion is ambiguous.

MT neurones - Provide the solution to this. Many MT cells respond to global motion = pattern cells. They can receive input from a series of component cells each of which has a different preferred direction but all of which have RFs covering the same area and integrate this to respond to global motion.

n.b. MT has a columnar architecture that is systematically organised with regards to the preferred direction of motion.

Question 30

How can light be measured?

How is luminance different to brightness?

Answer 30

a) Radiant energy, photons per unit time, candela /m squared (luminance)

b) Luminance - the amount of light emitted or reflected that can be measured.
Brightness - the subjective attribute of light - what we perceive

Luminance does not = brightness. Latter is affected by the relative absorptions of nearby areas. therefore may perceive 2 equally luminant areas to be different brightness.

Question 31

What is wavelength discrimination?

Why does it require >1 different spectrally tuned photoreceptor?

How did Von Frisch demonstrate this?

Answer 31

a) ability to detect changes in wavelength independently from changes in intensity = a pre-requisite for colour vision.
b) due to univariance - single spectrally tuned receptor is colour blind. > 1 needed to disentangle wavelength from intensity.
c) Demonstrated colour vision in honey bees, using a grid with 1 blue square and the rest being different shades of grey. Bees were able to distinguish the coloured square, showing chromatic sensitivity.

Question 32

What are the differences between rods and cones?

Answer 32

More cones than rods
Rods - scotopic, cones - photopic 
Rods - rhodopsin, cones - cone opsin
Rods - high sensitivity
Rods - can get saturated 
Rods - lower acuity
Rods - slow to respond

Question 33

Patient case examples for:

a) Akinetopsia - inability to detect motion
b) Visual agnosia - inability to recognise objects

Answer 33

a) Patient LM
- Lady had a stroke - bilateral damage in area MT - highly specific and selective loss of visual motion. She could still perceive some forms of motion, which was better centrally. Perhaps MT is more involved in the periphery?

b) Patient DF
- Carbon monoxide poisoning - bilateral occipitotemporal damage - impaired ability to recognise objects.She could scale her grip to the dimensions of the object but couldn’t perceive those dimensions. - unable to estimate dimensions until it came to picking objects up.

Question 34

In what layers do the 3 parallel systems terminate in the PVC?

Describe the receptive field of cortical cells.

Answer 34

a) Magnocellular in 4C alpha
Parvocellular in 4C beta
Koniocellular in blobs strung along the centres of ocular dominance columns in layers 2 and 3.

b) Most are orientation tuned and elongate.
Many are direction tuned.
They are different sizes to allow them to respond accordingly.
Some have separate on and off zones (simple cells) whilst others have mixed regions - spatially homogenous (complex cells)

Question 35

Describe the layered structure of the LGN.

Answer 35

6 layers: inner 2 are magnocellular, outer 4 are parvocellular. Koniocellular inputs go to inter laminar layers.

Layers 2, 3 and 5 receive input from the ipsilateral eye.

Layers 1, 4 and 6 receive inputs from the contralateral eye.

Cheng et al 2008

Question 36

Which neurotransmitters are involved in the;

a) on-cone pathway?
b) off-cone pathway?

Answer 36

a) Reduced illumination - increase in glutamate release - acts at MGLUr6 - produces a hyperpolarisation - weird that its inhibitory.

Illumination - glutamate release - inotropic receptors on RGCs - AP firing in light.

b) Reduced illumination - glutamate release - depolarisation - normal for glutamate (excitatory) - AP firing in dark.