L5&6 The Visual System

Question 1

What is light ?

Answer 1

The basis of visual perception
A wave of electromagnetic energy
Amplitude is the depth of the wave
Wavelength is the length of the wave
We’re only sensitive to a narrow range (380-750 nanometers)

Question 2

What are degrees of visual angles ?

Answer 2

Visual field - the outside world that the retina can see
Degrees of visual angle - in vision science, we like to refer to the size of an object in degrees of visual angle

Question 3

What is the anatomy of the eye ?

Answer 3

Anterior chamber - front of eye, filled with fluid
Cornea - front of eye, transparent
Pupil - let’s light enter
Iris - coloured muscle, stretches
Ciliary body - changes lens shape
Vitreous chamber - jelly liquid in the eye making it spherical
Retina - where photoreceptors are at the back of eye
Fovea - light ends here, looks for detail
Choroid layer - light absorbers
Sclera - membrane behind the eye
Optic nerve - absorbs nutrients
Optic disk - no photoreceptors (blind spot)

Question 4

What are cones ?

Answer 4

Concentrated near the fovea
6 million per eye
Sense colour
High resolution
1 to 1 with ganglion cells

Question 5

What are rods ?

Answer 5

Concentrated away from the fovea
120 million per eye
Monochrome
Low resolution
Many to 1 with ganglion cells

Question 6

What is the blind spot ?

Answer 6

Optic nerve and retinal artery enter the eye above the retina, creating a physiological blind spot
The eye fills in the missing information with whatever it can

Question 7

What is the duplex retina ?

Answer 7

It tries to catch photons and signal the presence of light
It must be able to operate in all the different conditions we’re likely to encounter
1. Cone driven, photopic system, high acuity, low sensitivity
2. Rod driven, scotopic system, low activity, high sensitivity
Can’t perceive colour when it is dark

Question 8

What is a retinal ganglion cell ?

Answer 8

1.25 million RGC
Summaries information by using single cell recordings

Question 9

What is a single cell recording ?

Answer 9

Place an electrode next to the axon of an RGC to record electrical changes in the axon
Move the position of light around until we influence RGC activity - this is the receptive field

Question 10

What are receptive fields ?

Answer 10

RGC action potential firing pattern is influenced by the light in their receptive field
They have a baseline firing rate (number of action potentials in a second)
This can go up or down
They have a centre region

Question 11

What are centre-surround antagonism ?

Answer 11

An on centre cell
If light is turned on in the centre region, the RGC firing rate will increase
If a light is turned on in the surround region, the firing rate will decrease
An off centre RGC has opposite firing
Light outside of the RF will not influence activity

Question 12

How do receptive field size change ?

Answer 12

RF are smallest in the fovea (0.01mm)
Providing high spatial resolution
10mm from the fovea, RF sizes increase by a factor of 50
Provide low spatial resolution (but good light sensitivity)
Further away from the gaze, the less detail

Question 13

What is the purpose of the centre surround organisation ?

Answer 13

Identifies edges in an image
Objects can be distinguished from the background by sudden changes in reflected light
If an edge is positioned appropriately, there will be a response from the cell
Each RF will respond optimally to a bar of a particular width depending on its size

Question 14

What is the visual pathway ?

Answer 14

Retinal ganglion cells produce APs
These project visual information through the optic nerve
They switch sides at the optic chiasm to go to the contralateral side of the brain
The optic tract projects to the lateral ganiculus nucleus in the thalamus
Optic radiations project to the visual cortex

Question 15

Why do nerve fibres from each eye cross ?

Answer 15

To allow for visual hemifield-specific processing in later visual areas
Visual information on the left will fall on the temporal retina of the right eye and the nasal retina of the left eye

Question 16

What is the lateral geniculate nucleus ?

Answer 16

Part of the thalamus
An important relay station with the brain
It has 6 layers:
Cells 1 and 2 are larger and get input from M cells (magnocellular layers)
Cells 3-6 are smaller and receive input from P cells (parvocellular layers)
Cells with input from K cells are sandwiched between the magnocellular and parvocellular layers

Question 17

What are the lateral geniculate nucleus connections ?

Answer 17

Ipsilateral fibres (not crossed) - input to the LGN in layers 2(M), 3(P), 5(P)
Contralateral fibres (crossed) - input into layers 1(M), 4(P), 6(P)
Each eye provides contralateral input to one LGN and ipsilateral input to the other

Question 18

What is retinotopy ?

Answer 18

An orderly map of the retina
Adjacent areas of the retina are represented in adjacent regions in the LGN
The spatial relations of the retina are therefore preserved (retinotopic map)
Each layer has a map
LGN cells are therefore sensitive to specific regions of visual space

Question 19

What are some characteristics of LGN cells ?

Answer 19

Colour - nearly all P cells are colour sensitive, this is the basis of colour opponency
M cells respond to all colours

Question 20

What is the structure of the visual cortex ?

Answer 20

LGN projects to the primary visual cortex
V1 has 100m cells per hemisphere and organised in layers
LGN input comes into V1 at layer 4

Question 21

What are ocular dominance columns ?

Answer 21

Cells in layer 4 are driven by the input from one eye only
If cells receive input from the right eye, cells above and below it will also receive input
Adjacent blocks of cells either side will receive inout from the opposite eye
This creates a pattern of ocular dominance

Question 22

What is cortex retinotopy ?

Answer 22

Adjacent regions of the retina are mapped onto adjacent regions of the cortex
Distributions of cells associated with each retinal region is distorted
80% of the cortical cells are devoted to the centre of the visual field

Question 23

What is cortical magnification ?

Answer 23

The disproportionate weighing of cortical power
Mirroring how the vast majority of RGC are devoted to the fovea

Question 25

What is orientation selectivity ?

Answer 25

Unlike RGC and LGN, most cortical cells have marked preference for particular orientation
Cortical RFs are organised and shaped differently to obtain maximum response
Staining can tell us about the orientation preference in cortical cells

Question 26

How is size and location effected ?

Answer 26

Cortical cells come in different types
Some are sensitive to location and some are sensitive to size
Hypercomplex cell - optimum response depends not only on orientation but also contour length, when the bar length matches the length of the receptive field

Question 27

What is binocularity ?

Answer 27

Cells in V1 layer 4 are monocular
Other layers are binocular meaning they can be driven by either eye
If the same stimulus is delivered to other eyes, the response is more vigorous
binocular cells have 2 RFs and they are matched in type

Question 28

What are colour sensitive cells ?

Answer 28

Concentrated in the cortical blobs
Each blob is centred on an ocular dominance column
Cells will either have blue/yellow or red/green opponency (not in the same blob)
They receive their input from lower layer 4 which gets their input from parvocellular LGN layers
No preference for a particular orientation

Question 29

What is direction selectivity ?

Answer 29

A large proportion of cortical cells display preferences for moving stimuli in a particular direction
Motion sensitive cells usually respond to only one direction - known as direction selectivity
Simple cells - slow motion
Complex cells - fast motion

Question 30

What are columns and hypercollumns ?

Answer 30

Visual cortex is composed of column cells
Each column consists of cells with the same orientation and ocular dominance preference
18-20 columns transverse a complete range of orientations and ocular dominance
A collection of adjacent columns is called a hypercollumn (ice cube model)
Every HC contains the neural machinery required to analyse multiple attributes of an image