Vision ll Flashcards
retinal ganglion cells project to many places
~90% of retinal projection is to the lateral geniculate nucleus
and it is also projected to pretectum:reflexive eye movements and pupil size.
lateral geniculate nucleus
LGN is a nucleus (cluster of neurons) in the thalamus
~90% of retinal projections are to the LGN, which subsequently sends significant projections to the cortex.
Cortex is associated with “conscious” vision.
Why do you want binocular vision
– if you’ve seen a 3D movie you know – it lets you easily judge relative depth and perceive distance.
So primary visual cortex is the first time that you get converging information from the two eyes – binocular cells
The image on the right (nasal) portion of the left eye projects to the
right side of the brain
- decussation of the optic fiber
the image on the right (temporal) portion of the right eye, projects to
the right side of the brain
the image projected to the temporal portion of each eye do not cross at the optic chiasma, it goes to that side of the brain where the eye is located
Contralateral vs Ipsilateral
Contralateral – on the opposite side of the body (used for left-vs-right)
Ipsilateral – on the same side
Decussation vs Partial decussation
Decussation – crossing of the axons from one side of the body to the other
Partial decussation – when only some of the axons in a nerve cross from one side to the other
Partial decussation – each eye receives information both sides of the visual world.
However, each side of the brain only receives information from one side of the visual world.
Fibres from each nasal hemiretina
decussate in the
optic chiasm:
The left visual cortex represents the right visual field.
The right visual cortex represents the left visual field.
Transection of optic nerve
- Functionally equivalent to closing one eye
left optic nerve cut => monocular vision, only see with right eye – but both sides
Transection of optic tract (or LGN or V1)
Only see things in one hemifield
Both eyes still functional
left optic tract cut => retain binocular vision – but only for left visual hemifield.
Transection of optic chiasm
Only crossing fibres are affected
Both eyes still functional
Lose binocular stereoscopic vision
Bitemporal hemianopia (e.g. due to pituitary tumour)
optic chiasm cut=> only central vision. Remember, nasal portion of retina – which “sees” temporal visual field decussates. This is lost, so we are left with central vision.
For bonus points – processing of each part of the visual field is monocular, because only get R visual field from L eye and L visual field from R eye.
what happens if pituitary gland swells?
Optic tract & chiasm wrap around pituitary gland, so swelling of the pituitary can block nerve transmission
what are two main visual “streams?
Parietal / dorsal / where pathway
Temporal / ventral / what pathway
Role of Parietal / dorsal / where pathway
cortical areas are specialised for processing object position and motion
vision for action / interacting with environment
Temporal / ventral / what pathway
- Cortical areas are specialised for processing object form and identification
- vision for perception
what is Brodmann areas – anatomical classification (cytoarchitecture
Brodmann areas – anatomical classification (cytoarchitecture
discrete regions of the brain could be distinguished based purely on anatomical criteria:
the density of cell bodies in different layers
The patterns of myelination
In these pictures, each anatomically defined cortical area is shaded differently.
How are they connected?
Serial vs Parallel processing
Many computations are carried out simultaneously (in parallel)
e.g. Photoreceptors all work in parallel; MT and V4 act in parallel
Some processing increases in complexity and must occur serially
e. g. in the ventral stream
- V1 contains neurons that encode orientation
- V4 contains neurons that encode simple shapes
- TE contains neurons that encode objects and faces
Different ganglion cell types tile the retina:
parallel processing with functional segregation
>12 parallel circuits, with unique classes of ganglion cell
Each circuit receives inputs from the same cone photoreceptors, but the inputs are processed in different ways.
Parasol cells (M-type cells)
- large cell bodies, dendritic arbors, receptive fields
sensitive to rapidly changing stimuli
not colour sensitive (inputs from all cone types)
project to Magnocellular LGN layers
Midget cells (P-type cells)
- small cell bodies, dendritic arbors, receptive fields
sensitive to fine stimulus features
colour sensitive (selective cone inputs)
project to Parvocellular LGN layers
predominantly in fovea
Size of cell body reflects function
Midget cells are able to detect fine spatial features because they don’t integrate across as many photoreceptors
parasol cells – sensitive to lower contrasts – larger regions of space
• Larger cells => faster conduction velocities / higher temporal resolution.
•Intrinsically photosensitive RG cells – project to suprachiasmatic nucleus – circadian rhythms; also control pupillary light reflex.
• Prevailing theory
– Parvo = object recognition / shape; Magno = motion.
Major retinal projection is to the
lateral geniculate nucleus (LGN) in the thalamus
The six layers of the LGN each represent the contralateral visual field.
Roughly speaking, receptive field properties of neurons in LGN are similar to those in the retina.
LGN receives more inputs from cortex than from retina – feedback is clearly important.
•LGN is first region where attention can influence sensory processing.
LGN is a six-layered structure: parallel processing with functional segregation (again)
Parvocellular Layers (3-6)---Inputs from midget RGC Magnocellular Layers (1 & 2)----Inputs from parasol RGC there are also intermediate, or Koniocellular layers between each layers
where do each layer receives information from?
Each layer is retinotopically organised and has a complete representation of the contralateral visual field.
Layers 1, 4, 6 = receive inputs from contralateral eye. Layers 2, 3, 5 = receive inputs from ipsilateral eye.
If you drive an electrode perpendicular to the layers, you will encounter neurons with the same spatial receptive fields.
• Magno – low contrast, high temporal acuity, poor spatial acuity, monochromatic
• Parvo – colour, fine spatial acuity
The major projection from the LGN is to
Primary visual cortex = V1 = striate cortex = area 17
functional segregation in V1
Functional segregation evident in retina & LGN is also evident in V1.
M-cells project to layer IVCα; P-cells project to layer IVCβ.
Complex connectivity then occurs within a column of neurons. (long axon connection )
Outputs from V1 are separated into two streams
- Parietal / dorsal / where pathway
- Temporal / ventral / what pathway
Retinotopic maps in V1
Adjacent neurons in V1 respond to stimuli in adjacent regions of visual field
Receptive field locations are determined by their inputs (photoreceptors => RGC => LGN => V1)
- Foveal over-representation (more spatial area = more neurons = higher acuity)
Is Spatiotopic organisation is common in sensory systems ? explain
YES!– in somatosensory cortex, adjacent areas of skin are associated with adjacent regions of cortex. Similarly in auditory system. In visual system, adjacent regions of cortex are associated with stimuli that are close in visual space.
The terminology “simple” and “complex” comes from Hubel and Wiesel.
Simple and complex cells:
- respond best to elongated bars or edges.
- are orientation selective
Simple cells:
have spatially segregated ON and OFF subregions
position of the bar within the RF is important
are often monocular (i.e. only respond to inputs from one eye)
Complex cells:
have spatially homogeneous receptive fields (i.e. no segregation of ON/OFF subregions).
position of the bar within the RF is unimportant
are nearly all binocular.
What stimulus is most likely to cause an AP in the simple cell?
What stimulus is most likely to cause an AP in the simple cell – something that generates simultaneous APs in the LGN cells, leading to simultaneous EPSPs in the simple cell.
•Working backwards, this requires a particular pattern of visual stimuli to affect the photorecteptors … RGC etc.
•This highlights the benefits of center-surround receptive field organisation – it’s optimised for encoding edges, borders, and enhancing contrast. It’s also computationally efficient.
Putative circuits to explain “Simple cells” and “Complex cells”
Theoretical model:Multiple LGN cells with collinear receptive fields synapse onto a single V1 simple cellto the stimulus that best activates all LGN neuron
The V1 simple cell’s preferred stimulus corresponds
A map of orientation in V1
1-Cells within the same cortical column have similar tuning properties (i.e. they prefer the same orientation and have similar spatial receptive fields)
2-Cells in adjacent columns prefer similar orientations.
3-The map for orientation exists on a coarser spatial scale than the map for position
If you drive an electrode perpendicular to cortical surface – neurons all have
same RF and same orientation preference.
If you drive an electrode tangential to cortical surface – neurons have
adjacent RF and slowly changing orientation preferences.
• Important for two reasons: suggests there are constraints on the way cortex is ‘wired” up during development.
• Secondly, it suggests that lines and their orientations are important, primitive features in the visual world.
Tuning in V1
• All neurons respond best to elongated edges and are orientation selective
• Neurons respond to specific regions of visual space
• Some neurons are dominated by parvocellular LGN inputs
=> sensitive to colour
• Some neurons are dominated by magnocellular LGN inputs
=> motion-sensitive
• Within a V1 cortical column, neurons have the same spatial receptive field and prefer the same orientation
• Moving across adjacent V1 cortical columns, neurons systematically change their spatial RF location and preferred orientation
Neurons in middle temporal area
underlie motion perception
Individual MT neurons are tuned for direction
Groups of MT neurons show diverse tuning
In humans, MT lesions impair motion perception (akinetopsia)
In macaque monkeys, lesions impair eye movements
Experimental microstimulation biases motion perception
Perception of ambiguous stimuli is predicted from MT activity
Neurons in inferotemporal cortex underlie
face perception
Individual IT neurons are tuned for direction
Groups of IT neurons show diverse tuning
In humans, IT lesions impair face perception (prosopagnosia)
Experimental microstimulation biases face perception