V1 to MT - evidence for convergence

Question 1

Where do the cells that innervate MT come from? What are the types of these cells? Where are these located? Correlates with Harvard Hypothesis?

Answer 1

4B. Circles and triangles. Some aligned under blobs, some underblobs. No correlation. HH would expect interblob to be devoid of MT cells.

Question 2

Explain Nassi and Callaway’s experiment.

Answer 2

Take out brain, cut into slices. Bathe in caged Glutamate. Light of particular wavelength uncages. Magno(4CA)/Parvo(4CB) recipient cells can be driven. Most axons in V1 go up.

Question 3

What is observed in 4Calpha in Nassi&Callaway?

Answer 3

Neurons that send to L3 and then outside V1 respond. Also inside though. Pyramidal cells.

Question 4

What is observed in 4Cbeta in Nassi&Callaway?

Answer 4

Only pyramidal cells that send to layer 3 and not outside of V1 respond.

Question 5

What is observed in 4B in Nassi&Callaway?

Answer 5

Some pyramidal (long dendrite) some stellate. Some of both go to V2 (where in V2 depends on under blob/inter). Others send directly to MT.

Question 6

Where do neurons in 4B send axons to? What types of neurons are there in this layer? What does each type do?

Answer 6

They send to MT and V2. Stellate and pyramidal both send to both. Nassi&Callaway.

Question 7

V1>V2 neurons, where do they get their input from?

Answer 7

4B>V2 mixture of magno-dominated stellate cells (20%) and mixed input stellate cells (805).

Question 8

V1>MT neurons, where do they get their input from?

Answer 8

80% magno-dominated stellate and pyramidal cells driven by 4CA.

Question 9

Where do blobs/innerblobs get input from?

Answer 9

4CB almost exclusively to innerblob. 4CA almost exclusively to blobs. Blobs get konio input too.

Question 10

What do stellate cells of 4B do to MT?

Answer 10

They provide MT with monocular directional input that is magno driven. Selectivity for mvmt of stimulus direction. Coarse. MT sharpens this.

Question 11

What do pyramidal cells of 4B do to MT?

Answer 11

They provide MT with binocular disparity-tuned signal through a relay in V2.

Question 12

What’s the summary of the path from LGN to MT?

Answer 12

Where you get your input and where you send your axons to is tightly coupled. Magno>Py/St. Parvo>Py>V2>MT. St>MT.

Question 13

What type of cells send axons out of V1?

Answer 13

Pyramidal mostly. A few spiny stellate. Some pyramidal stay local though.

Question 14

What’s the fundamental cortical circuit?

Answer 14

LGN>4C>4B/3/2. They drop off collateral at 5. 2-4B>extrastriate. 4C>6>4C. LGN>6. The L6 is modularory.

Question 15

What’s the property of neurons in 4A? What area has the opposite property?

Answer 15

They receive koniocellular input. 4A blue-OFF. Blobs have blue-ON.

Question 16

What’s the Peters rule?

Answer 16

Any cell that can receive an excitatory synapse will receive thalamic input if it’s in layer 4. TLDR: Connections to thalamus in cortex are stochastic.

Question 17

Explain how layer 4C connects to layer 3.

Answer 17

They are very specifically connected. 4CB interblob 4CA blob.

Question 18

Explain the salient feature of orientation columns.

Answer 18

pinwheel. There are places where all orientations come together. Pinwheels are mostly at the center of ocular dominance columns. Orientation columns intersect ocular columns at right angles. Precision at center of pinwheel is very high because you can move a little bit and find a cell tuned to a different orientation.

Question 19

Do pinwheels align with blobs?

Answer 19

Both are at the center of coular dominance columns. Compare orientation selectivity to blob location. Half the time a PW center will occupy the same location as a blob. 50% not bad, blobs are 15% of layer 3. this means it’s better than chance.

Question 20

What are simple cells?

Answer 20

P/M cells are mostly center on, surround off. 7 LGN neurons send their axons send their axons to a cluster of 160 cortical cells. They have the strongest response that orient with those neurons. A convergence of excitation model. Simple cells get their orientation selectivity from arrangements of RF of LGN neurons that innervate them. You can use spots of light to determine RF of a simple cell.

Question 21

What are complex cells?

Answer 21

Complex cell: several simple cells converge to a complex cell. RF is bigger. Can’t map with spots of light. If you move a stimulus, you get a nice response in one direction, but not response in another direction.
Have orientation selection because the simple cells that innervate them have orientation selection.

Question 22

What’s wrong with the complex cell model?

Answer 22

If you increase contrast, the response of the V1 gets better. Broadness of tuning in LGN and retina increases as you increase contrasts. BUT orientation selectivity of simple cells does not respond to contrast. If you increase contrast, increase drive of RGC in LGN neurons, but you don’t see increase in simple. So there’s sth wrong with model.

Question 23

Explain the hypercomplex property.

Answer 23

Take a stimulus, make the bar longer than the RF, but some have a hypercomplex property. If you make it much longer than the RF, the response drops, becomes excitory and inhibitory.

Question 24

What are the models for orientation selectivity?

Answer 24

1. Gain control model: Not only effect of excitatory but also inhibitory. After 50 ms orientation tuning becomes tighter. Enters cortex poorly, cortical circuit tunes it. Gabor.
2. Membrane potential thresholding for tuning out weak inputs.
3. Different orientation tuned excitation neurons all send to a GABA neuron which sends back to them inhibiting broad angles.