Neural Processing of Vision II : LGN to Visual Cortex

key 2 points to take away from the visual pathway

• Information from right visual field goes to left visual cortex and vice versa

• nasal fibers will cross but temporal fibers won’t in order for that to work ^^^

• extra: Info that hits the temporal retina comes down optic nerve and stays on same side of optic chiasm – if info hit nasal retina those fibers go out optic nerve and optic chiasm and go to LGN and V1

sketch the visual pathway from retina to cortex

• both eyes, the right SC, the right LGN, the right V1

• just think of it being the opposite - so if information is coming from L VF then it will go the RIGHT and VV

where do the optic nerve fibers from the retinal ganglion cells go - visual pathway

• Lateral geniculate nuclei (LGN) — 90% terminate here in macaque.

• Superior colliculi— 10% terminate here in macque

some, but very few may also go to:

• Suprachiasmatic nuclei—diurnal rhythm

• Pretectum—pupil diameter

• Pregeniculate—may be diurnal rhythm, unclear

• Accessory optic system—maybe gaze stabilization, unclear

explain basic function of the superior colliculus

• play a role in integrating sensory information (visual, auditory, somatosensory) into motor signals that help orient the head toward various stimuli and lead to action

• Has vision, sound and touch – integrates across these 3

explain structure of the superior colliculus (4)

• Layered structure (7 layers): top 3 layers have visual input, lower layers have auditory and somatosensory input

• Visual input comes from retina AND V1, V2, V3, V4 and MT.

• Cells are retinotopically organised (the brain keeps the same spatial layout as the retina so it can interpret visual scenes accurately, like a mental map of the image)

• Visual cells in superficial/intermediate layers have ill-defined ON and OFF regions responding to just about any visual stimulus eg. bars of any orientation, spots, light flashes (so concerned not so much with “what” an object is, but “where” it is) - Not good at analyzing form but

  good for knowing where something is

importance of the superior colliculus / role (2)

• SC has an important role in visual orienting reflexes

• SC is important for control of visually guided saccadic eye movements (frontal eye fields also important)

explain the basic structure of the LGN - layers

• •The LGN has a very distinct, layered appearance (in humans it has 6 layers)

• geniculate “with bent knee” – appearance

• Layers 1 and 2 – larger cells – magnocellular layers

• Layers 3,4,5,6 – smaller cells – parvocellular layers

• Inter-layers – even smaller cells – koniocells

  magnus “large”; parvus “small”; konis “dust”

• Info from 2 eyes is kept separate in distinct layers

• Fibres strictly segregated at LGN

explain projection of rGCs to the LGN

• P rGCs project to parvocellular layers of LGN

• M rGCs project to magnocellular layers

• K rGCs project to very small koniocells between the major layers

explain how the layers of the LGN are said to be “in register” / retinotopic organisation

• each layer contains an orderly representation or map of the retina (a retinotopic map)

• so that axons carrying information from adjacent parts of retina (visual field) connect to adjacent geniculate cells

explain difference and similarity between the LGN cells and rGCs in terms of receptive fields

• Similar to rGCs, LGN cells have approximately circular receptive fields and centre and surround components interact antagonistically (ON- and OFF-centre cells)

• surround of LGN receptive field exerts stronger inhibitory effect on centre than it does at the rGC level – highlights difference between light and dark slightly

• Thus LGN cells serve to amplify differences in illumination i.e. contrast and edges

explain the differences between the cells in the layers of the LGN

• Differences between cells in parvocellular and magnocellular layers are similar to those initiated in P cells and M cells of the retina.

• Functions of Parvo and Magno cells clearly segregated at LGN, suggesting that they serve different roles in serving visual perception

describe the functional characteristics of the parvo cells in the LGN

• Nearly all are differentially sensitive to colour of light imaged in RFs – they are “colour opponent” eg. +R/-G or +G/-R

• RFs are 2-3 times smaller than magno cells at any particular eccentricity; important for analysing spatial detail

describe the functional characteristics of the magno cells in the LGN

• Cells not differentially affected by colour; response depends on relative intensity of centre and surround.

• Respond vigorously to fast abrupt fluctuations in light intensity within their RFs

• larger RFs

describe the functional characteristics of the konio cells in the LGN

• occupy interlayer regions of LGN

• in LGN, they have relative large RFs compared to those for P cells and M cells (but very small RF centres)

• they show selectivity for B versus Y, colour processing

• there are afferents from SC and some respond to sound and touch as well as visual stimulation

what can receptive fields at the level of the LGN code for

can code for size AND colour

what is the function of the LGN

• Pre process information before it is sent to the cortex – enables a more efficient onward journey to V1

• But also a receiving area from V1 to let the LGN know when it is ready for more signals or not - visual cortical feedback accounts for 30% of synapses

• modulate incoming information, integrating across large retinal areas and emphasizing sudden changes in stimulation

• Important regulatory function - regulating strength of signals sent to cortex and pre-sorting information into significant streams

explain what the V1 one is - basic

• Area V1 (striate cortex) is MAIN receiving area for LGN neurons

• There are at least 30 distinct visual areas beyond V1 – called extra-striate cortex

describe the layers of V1 (4)

• Anatomically split into 6 main layers (numbered 1-6), though layer 4 is subdivided into 4a, 4b and 4c

• layer 4c further subdivided into 4ca and 4cb

• Most of input arrives in layer 4c – magnocellular pathway terminates in 4ca; parvocellular pathway terminates in 4cb !!! - need to know this

• K cells terminate in layers 2, 3 and 4a

explain retinotopic mapping/organisation in V1 compared to in other structures (LGN retina SC and V1 all retinotopically mapped) (3)

• the layout of the retina is preserved in the primary visual cortex (V1) — like a visual "map" of what your eyes see

• Each point on the retina (where light lands) connects to a specific point in V1

• difference with V1 is cortical magnification - the central part of your vision (the fovea) is given much more space in the brain than the peripheral parts - central information is magnified - to allow for high-resolution vision

explain the main differences in the RFs at the level of V1 compared to LGN

• RFs are NOT circular in V1 - elongated

• can now code for orientation at V1 due to the elongation of the RFs

explain orientation selectivity in RFs in V1 (4)

• Any particular cell will respond only if orientation of edge or line falls somewhere within rather narrow range – about 15 deg to either side of optimal

• Cells in layer 4c, less orientation selective than those in other layers

• Put stripes that match to its preferred orientation – fires a lot // If don’t – wont fire at all – if 90 degrees off

• Orientation sensitivity is strong

what are receptive field profiles in V1 split into - 3

3 types of cells: simple cells, complex cells and hyper complex cells

explain characteristics of simple cells (5)

• Have elongated/not circular RFs with elongated excitatory and inhibitory regions

• Respond best to a) a bar or b) an edge of correct orientation and size in correct position on retina

• Thus selective for orientation, size and position

• Display little or no spontaneous discharge

• well defined ON and OFF regions

explain characteristics of complex cells (5)

• Have larger RFs and do not exhibit distinct ON and OFF regions - cannot be mapped out

• Respond best to rapidly flickering or moving stimuli - Exhibit sensitivity to motion and direction

• have a preferred orientation and size like simple cells, but are not so specific for stimulus position - Can respond to stimuli regardless of location within their receptive field

• Respond well to edges and bars of light but are less selective for position

• Because their responses are not simply given by linear summation of sensitivities of subregions, they are described as non-linear

explain the characteristics of hyper complex cells / end stopped cells

• Responses depend not only on contour orientation but contour length as well - line also has to be a certain length for it to fire a lot

• Maximum response is to bar whose length and width “fits” the RF – extending the bar length beyond this optimal value dramatically reduces cells response

• End-stopping found amongst both simple and complex cells – though originally thought to be property only of hypercomplex cells

summary of simple, complex and hyper complex cells

Cell Type	Orientation Selective?	Position Sensitive?	Motion Sensitive?	End-Sensitive?

Simple

✅ Yes

✅ Yes

❌ Not much

❌ No

Complex

✅ Yes

🚫 Less

✅ Often

❌ No

Hypercomplex

✅ Yes

🚫 Less

✅ Often

✅ Yes

explain orientation selectivity in terms of the oblique effect ()

• If you measure the orientation selectivity of every cell in v1 - MORE cells are biased to H and V orientations than obliques

• = higher sensitivity for H and V than obliques

explain binocularity in cortical RFs - first time we see binocularity in the visual pathway

• information comes to 4c separately from RE and LE layers from the LGN - once it leaves 4c ANY cell will respond to BOTH the RE AND LE

• Once leave layer 4, ocular segregation gives way to binocular integration – with few exceptions, all cells outside layer 4 are binocularly driven

in total what can we code for at V1 - RFs

• Coding for direction of motion develops in V1 also – by combining 2 cells

• so we can calculate direction of motion – as well as colour and size and orientation

explain the organisation of V1

• orientation: columnar organization of preferred stimulus orientation – all cortical cells in each column have similar preferred orientation – orientation column

• ocular dominance: columnar organisation of preferred eye or ocular dominance columns (remember most cortical cells are driven by both eyes)

• colour:

RFs BEYOND layer 4c can code for what…

• colour, size, orientation, direction of motion and binocular vision

summary of visual pathway journey

• rGCs synpase to LGN (10% to SC) = P to parvo, M to magno and K to konio – synapses coming back from VC TO LGN

• LGn synpase to VC – V1

Describe what hypercolumns are

Full range of orientation - both eyes are represented- aggregation of adjacent columns