Vision- from retina to cortex Flashcards
“The problem of vision”
What is there in the visual signals our eyes receive?
There is inherent ambiguity eg. in shape
What are some examples of ambiguity?
Cars presented at different viewpoints but its the same car. This will pass different shapes onto your retina. You might see different features as well. But we figure out these viewpoint and conclude its the same car.
You can have 2 different shapes that cast the same shape onto the retina. Due to the way light and perspective works, those 2 shapes (although different in their orientation and identity) might be cast onto the retina the same shape at the back of the eye.
Spinning dancer is also an example of ambiguity
List structures of the eye and their function
Pupil: where light enters eye (it passes through the lens)
Iris: (coloured part around pupil) adjustable aperture, constricts in bright light to make pupil smaller. The aperture determines how much light enters the eye.
Cornea and lens: focuses light on retina
Accommodation: ciliary muscles change shape of lens to bring objects into focus at different distances
List other structures of the eye and their function
Lens: helps focus light onto the retina
Ciliary muscles: allow for accommodation - when focusing on something change the shape of the lens
Retina: red layer which runs along the back of the eye
Fovea: high definition vision
What are photoreceptors? + photo transduction
Cells with light sensitive photopigments in outer segments
Photo transduction- production of AP so nerves begin to fire
What are the 2 broad classes of photoreceptors?
Rods:
* Contain rhodopsin, respond in dim light (generally used at night), none in fovea
* Features of rods- colour blind in rod vision
Cones:
* Split into three types with photopigments sensitive to different wavebands (long, medium, short) - daytime vision
* Daytime vision and colour vision
How does light enter the retina?
Due to the way our retina is organised, light comes in through the bottom and passes through all the neural machinery before it reaches the parts of the eye that are sensitive to light.
When staring at a dot on a black and white figure, what is happening to the photoreceptors?
Whilst staring at the dot, the photoreceptors on the parts of your retina which are underneath the light areas are adapting. They’re firing because light is landing on them, but they’re going to gradually reduce the amount of activity they produce (this is called adaptation).
- where do Retinal Ganglion Cells sit?
- what stage are the of retinal processing?
- downstream of photoreceptors
- last stage of retinal processing
Retinal Ganglion Cells:
What are the cells and what do they have?
Large Parasol ganglion cells: they have large receptor fields and connect to large areas of retina
Small Midget ganglion cells: smaller receptor fields- connect to a small number of photoreceptors
Cells have ‘Receptive Fields’ – the part of the retina from which the ganglion cell receives input
What are the parasol and midget cells linked to?
The parasol cells are connected to the magnocellular
The midget ganglion cells are connected to the parvocellular system
Single cell recording form retinal ganglion cells (RGCs)
- no light- baseline activity
- if we fill the light- baseline activity again (no change)
- For lots of classes of retinal ganglion cells, they have an arrangement where if light hits the centre then their activity will increase. This means lots of AP’s will be sent out.
- If you fill the surrounding light, then the activity decreases below baseline.
- This is called a centre surround receptive field
- On centre- off surround
- Off centre- on surround
Whats the difference between On centre- off surround and Off centre- on surround?
On centre- off surround: theres excitation (activity increases) when light hits the middle and inhibition if light hits the outside
Lateral inhibition
Off centre- on surround: excitation when light hits the outside and inhibition when it hits the middle
Retinal ganglion processing:
List 3 features
1- Poor at spotting gradual change
2- Good at picking out sharp edges:
When you have multiple on centre-off surround cells which span the visual field, you’ll get signals from the ones that correspond to where the edges are and no signal from the ones where theres no edge. This is because when theres an edge, there will likely be a difference between the centre and surround of the receptor field- no edge the centre and surround will be more the same.
3- Filters the input for useful information
Retinal ganglion processing:
Gradual example
If you spread the on centre receptor field, the only place you get activity is where the edge is there.
If you do this to a naturalistic scene eg. in a parking lot, it points out all the edges.
Being filtered by receptive fields similar to those of the retinal ganglion cells. The edges are picked out, whereas areas of uniform brightness (whether light or dark) look the same mid-grey.
What helps us with recognising things?
Edges
Perceptual effects of retinal ganglion cells……
Hermann Grid? (Baumgartner, 1960)
If you focus on one intersection- you will find that the dot disappears. But in the periphery there will be dots between these intersections. It was thought this was due to on centre- off surround visual fields.
Theres a difference between the junctions and the bits in between in terms of how much inhibition there is and the result is that is that you perceive these darker spots.
BUT, is it really due to retinal ganglion cells?
Wiggly version makes illusion go away-
Making it wiggly makes the effect go away so it probably isn’t due to retinal ganglion cells.
What else can retinal ganglion cells can help you detect?
Contrast
Perceptual effects of retinal ganglion cells……
Simultaneous contrast
image- black square on left, white square on right with 2 small squares in both
The patch (small square) on the right looks relatively darker than the left. This is because on the left hand side, theres less inhibition from the surround due to less light coming from the surround.
Simultaneous contrast
However it’s not this simple, theres something more going on…
The 2nd and 3rd tiles are physically the same but the one on the right looks darker. The difference in the overall picture is the presence of the table.
Where is the Lateral Geniculate Nucleus? + feature
In thalamus- sensory gateway (all senses except smell go through it)
What does the LGN have?
Layers (which are folded over). These have different roles visually.
What are the 3 distinct pathways running through the LGN?
The parvocellular system exists in some of these layers. (the lighter stripes)
The magnocellular system go into the sublayers which are closer to the middle of the structure.
The Koniocellular cells exist in between the layers.
What are the 3 distinct pathways running through the LGN linked to?
Magnocellular cells: movement & flicker
Parvocellular cells: colour & detail
Koniocellular cells: Blue-yellow
Visual Cortex:
- how much of the cortex is dedicated to vision
- what type of visual cortex are we concerned with
- Up to or more than 50% of cortex dedicated to vision (visual cortex and beyond)
- Primary visual cortex / V1 / striate cortex (has stripes in it- discovered by staining)
Retinotopy
V1 contains a retinatopic map. Over the surface of the cortex in V1, there is a point by point representation of all the points in your visual field.
Cells in V1 and orientation finding in cats
Visual images to anaesthetised cat
Slipping glass into projector moving black dot produced a faint line which lead to cats reponding.
They though maybe cells in V1 are a bit like retinal ganglion cells (like to see spots in their visual field and not spots in surround)
Instead the cells in V1 they were measuring from were sensitive to lines and lines of particular orientations
They found orientation selectivity in V1.
Detailed investigation of visual cortex
Lines indicate the sending of APs
For different orientations of lines, you get different responses
Eg. vertical line shows to be very responsive
Experimental setup used by Hubel & Wiesel.
Across V1 there are pin wheels. There are cells all round which respond to all of the orientations.
This tells us, across V1’s retinatopic map, you have a representation of every point and orientation of every point
Detailed investigation of visual cortex:
Orientation selectivity
Making a V1 receptive field from LGN inputs. The elongated bar detector has an excitatory central strip flanked by 2 inhibitory regions. It will respond best to a bright bar, tilted to the left, on a dark background.
V1 simple cell- if you take the output from centre-surround receptors - some on centre and some off centre- you can get a rectangular receptive field which is responsive to a certain or orientation of line and it won’t be responsive to other orientations of line.
Role of development/critical periods
Blakemore & Cooper (1970): task and results
Procedure:
* Prevalence of neuron types shaped by environment experienced early on
* Kittens raised in striped tubes from birth, vertical or horizontal stripes
* 5 hours per day in tube
* 5 months later: no response to orientation not in tube
Results:
* cats don’t show any behavioural response to other orientations.
* Recorded from cells in visual cortex: no neurons that respond to orientation absent in tube. Cats show no response to any other orientations.
* Example of neural plasticity, but with a critical period - ‘use it or lose it’
Beyond V1 has 2 streams of processing: what are they?
Ventral (runs along the bottom) and dorsal (runs along the top)
Beyond V1: Streams of processing
What is the difference between dorsal and ventral?
1- Dorsal (where)- used to plan movements, it projects up to the motor system
2- Ventral (what)- objects, colour, memory
Colour (ventral) & motion (dorsal)
Video illusion- Suchow & Alvarez, 2011 (Curr. Biol.)
Coloured dots appear in a cloud around a central white fixation point. The dots change colour smoothly and quite rapidly so the perception is of a twinkling changing set of colours. Then the cloud of dots all rotate together around the fixation point. Provided the observer looks at the white fixation point, when the dots are in motion the colour change is much less salient, even though the dots are still changing colour at the same rate.
Effect happens because colour and motion are separated in terms of those 2 streams.
What happens to processing of some features?
Processing of some features (i.e. edges) begins at the retina, and gets more complex “higher” in the visual processing stream.
What is V1?
brain area/ part of the cortex/ primary visual cortex
What is meant by the phrase “retinotopic map” in this context?
There is a representation of each point in the visual field (or hemi-field) arranged so that neighbouring parts of the retina are represented by neighbouring parts of V1/ the cortex/ brain.
