Week 5 Cortical Organisation

Question 1

Recall 5 steps of feature detection model for visual processing? (Assumption that each cell is accumulation of abilities from previous cells)

Answer 1

1. Rods and cones extract light at different intensities
2. They are pulled to form retinal ganglion cells
3. RG cells feed into LGN cells
4. Multiple LGN cells form simple cells
5. Simple cells form complex cells

Question 2

How does each stage of cell get more specific and complex?

Answer 2

1. Rods and cones = sensitive to uniform light intensities /
2. Retinal ganglion & LGN cells = sensitive to differences in luminance across space regions
3. Simple cell = sensitive to specific orientation and luminance polarity
4. Complex cells = sensitive to motion and colour

Question 3

What are the 3 problems with the feature detection model?

Answer 3

1. Assumption that each cell is accumulation of abilities from previous cells)
   In reality, complex cells cannot actually detect luminance polarity and location on the visual field, meaning, you would not be able to tell location and luminance
2. You would quickly run out of cells representing each stimuli, and you would need multiple cells for each object to account for changes and differences
3. Difficult to encode new information (similar to assimilation and accommodation in schema theory)

Question 4

How are retinotopic and topographic mapping similar and when is it lost?

Answer 4

Just as spatial arrangement of the real world is mapped onto the retina called topographic mapping, The spatial arrangement of the retina is also maintained and preserved when mapped onto V1, called retinotopic mapping

As the image from V1 is processed through the visual cortex, V2, V3, V4, the topographic quality of the image is slowly lost

Question 5

What are orientation tuned columns?

Answer 5

V1 simple cells in columns perpendicular (90˚ angle to cortex surface) are tuned to the same orientation and fire are preferred orientation

Question 6

Monocular vs. ocular dominance colomns?

Answer 6

1. Monocular columns receive input from one visual field
2. Ocular dominance columns receive both left and right visual input and are therefore binocular, but may receive dominant input from either left or right visual fields

Question 7

What are hypercolumns?

Answer 7

Hypercolumns are orientation tuned columns that cover an entire range of orientations within a specific visual region, made from ocular dominance columns and simple cells

Hypercolumns form the basis for the sensation of depth perception and encode stimuli of any possible orientation

Question 8

What are blobs?

Answer 8

Blobs are regions of the visual cortex with high amounts of cytochrome oxidase enzyme which affects cell’s energy in energy demanding visual regions such as near colour sensitive center surround cells

Question 9

What is the range in orientation that demarks one orientation type cell from the next?

Answer 9

10˚ degrees is the approximately difference in preferred orientation from one column to the next column of simple cells

Question 10

If 10˚ degrees is preference orientation difference, how do we discriminate smaller orientations of less than 10˚ degrees?

Answer 10

Instead of looking at a single cell, local population responses enable us to discriminate finer orientations of less than 10˚ degrees, ie. changing in seconds on a clock

Question 11

We initially have high amounts of visual firing in brain when we look at a new stimulus, and over time there is a reduction in sensitivity. How does this explain the tilt after effect?

Answer 11

Adaption - we adapt to familiar stimuli

1. Vertical lines vertical tuned cells will respond, ie. population response will occur, but also the cells tuned to slightly left and right of vertical lines
2. Presenting lines tilted slightly left away from the vertical responses will result in a stronger adaptation to left-oriented cells, slight adaptation in vertical-oriented cells.
3. When vertical lines are presented again, the population response will be now shifted to the right because the population response is going to be skewed away from the reduced sensitivity of left-oriented cells due to adaptation.

Question 12

What is PSE?

Answer 12

Plotting the proportion of seen oriented to the right vs the orientation of the stimulus, ie.

when the stimulus is on the right, you see 100% of the stimulus, when tilted to the left, you have 0% of responses tilted to the right

When its vertical, neither left or right, you would expect 50% of responses tilted to the right

Question 13

What is the oblique effect and why does it occur?

Answer 13

Oblique Effect is our tendency to have greater sensitivity and acuity for vertical and horizontal stimuli compared to oblique/slanted lines

This is because we have more cells tuned to vertical and horizontal orientations compared to oblique.

This may be due to elements of an industrial environment, ie. more vertical and horizontal contours in the environment from buildings, poles, signs, boxes, furniture, etc.

Question 14

How could we test whether our preference for straight lines genetically determined or shaped by visual experience?

(animals and patch)

Answer 14

Experiments on new-born animals, expose it to no stimulation/make it blind/ deprive it of visual input and measure the tuning bandwidths and orientation preferences

Monkeys have the same orientation tuning columns as normal adults, ie support for genetics

Selective stimulation:
Patch one eye - only monocular cells sensitive to the other eye, only monocular cells become tuned to the other eye ( binocular cells are shaped, the development of these cells are sensitive to visual stimulation in a critical window, ie. 4 months)

Question 15

How can the tilt aftereffect can be used to determine the tuning bandwidth of cells?

Answer 15

The broader the bandwidth, the further away from vertical the adaptor can be and still affect the perception of vertical lines.