Brain and Vision

Purves et al., Chapter 12 (and bits of Chapters 11 and 24)
Note that the Washington University Medical School's Neuroscience Tutorial has good coverage on this topic:
Central Visual Pathway

Projection to Brain

Fig. 12.1
Overall Visual Projection
Eye -> LGN (Lateral geniculate nucleus, genu= knee, part of thalamus) -> striate cortex
Temporal retinal field = nasal visual field stays ipsilateral at chiasm
Nasal retinal field = temporal visual field crosses to contralateral side at chiasm
From LGN to striate cortex = area 17 = V1
Retinotopy (like somatotopic organization) is preserved
*Eye -> pretectum - pupil size (iris) and control of lens (accomodation)
Eye -> superior colliculus - eye and head movements (Chap. 19)
Eye -> hypothalamus - to regulate circadian rhythms (see, in chapter 27)

*Fig. 12.2
pupillary reflex
Pretectum -> Edinger-Westphal nucleus -> cranial nerve III->ciliary ganglion ->parasympathetic fiber.
Note connection to both ipsilateral and contralateral sides after pretectum, so pupillary reflex should be bilateral.
Kids, go ahead and try this.
Important test in neurology.

Fig. 12.13 A & B
Thalamus
Cells have center - surround receptive fields like ganglion cells
1, 4, 6 contralateral -- thus 2, 3, 5 ipsilateral

Fig. 12.15A
large and small retinal ganglion cells
Magnocellular - large receptive fields for processing movement - connect to LGN 1 & 2
Parvocellular cells connect to layers 3, 4, 5, & 6 and process color, also for acuity

Cortical processing

Landmark papers
DHHubel & TNWiesel, Receptive fields, binocular interaction and functional architecture in the cat's visual cortex, JPhysio, 160, 106-154, 1962
TNWiesel, DHHubel & DMKLam, Autoradiographic demonstration of ocular dominance columns in the monkey striate cortex by means of transneuronal transport, Brain Res 79, 273-279, 1974 (see also J NIH Res, 5, 61-67, 1993)
DHHubel & TNWiesel, Brain mechanisms of vision, Scientific American September 1979 (vol 241, #3), pp 150-162.

Tom Yin's home page, follow links, Simple cell is a good video to show how work was done

Fig. 12.8 A & B
cat (monkey) looks at screen, cell responds best to line at angle
Striate cortex - physiology and anatomy
Hubel & Wiesel share 1981 Nobel for "information processing in the visual sytem"

Fig. 12.11
there are vertical columns of preferred angle (just like in somatosensory system)
presumably, to prefer line at angle, cell receives inputs from from alligned center surround cells
these are called simple cells
complex cell - line at algle moving in direction
hypercomplex cells - line has end - corner
vertical electrode penetration gives cells with all the same preferred angle
an oblique penetration tracks different angles
Note that there are 6 layers of cells, IV has inputs from LGN

"Philosophical question" -- does processing get to more and more levels of complexity until you find "grandmother cells" which recognize, specifically, your grandmother's face?

Fig. (like 12.13 B & C)
experiment to determine ocular dominance columns (0.5 mm wide)
There are cortical cells with input from one eye, from the other eye, and, in between, from both eyes.
Binocularly driven cells should be necessary for stereopsis, the kind of depth perception which relies on the focussing of both eyes.

Fig. 12.14
Apparently, cells can preferentially respond to disparity from fixation
depth-perception
stereopsis

Even higher order visual processing

With all that color processing in the LGN, it seemed odd how far the work on the cortex got without any mention of color

Fig. 12.16A
V4 - color but not movement
MT (middle temporal) - direction of movement but not color

Fig. 12.18
parietal stream - spatial vision
temporal stream - object recognition

Development of visual connections

Fig. 24.3
If a radioactive amino acid is injected into one eye, labeled proteins cross synapses at LGN and mark ocular dominance columns in cortex; this is detected by microscopic autoradiography.
Binocular cells connect up correctly at first

Fig. 24.4
Then there is a sensitive (critical) period in the first few months of life during which patterned visual input from both eyes is necessary to maintain binocular input to cortical cells.
Thus early visual defects like cataract or strabismus (cross-eyes or lazy eye) need to be corrected right away.

Here are autoradiographs. A of normal visual cortex, nd B after monocular deprivation from 2 weeks to 18 months in monkey

Exam questions from 2005 - 2007 relating to this outline

How would you test whether a patient's contralateral connection from the pretectum to the Edinger-Westphal nucleus were disrupted?

light to one eye would not consstrict the contralateral pupil

Specifically, what crossed the LGN (lateral geniculate nucleus) synaptic cleft to allow Hubel to see the ocular dominance columns using autoradiography?

3H labeled protein

Where do the axons of the the temporal retina (nasal visual field) go at the chiasm?

to ipsilateral LGN

What is the word for the special type of "depth perception" mediated by the parallax from focussing both eyes (and neural interactions of binocular cells)?

stereopsis

If a child were born with a congenital cataract in one eye, why should this deserve immediate attention?

brain connections from that eye would vanish without patterened vision

Light stimulation to one eye activates that eye's optic nerve. By what mechanism would there be efferent output to the iris from both occulomotor nerves?

connection to one pretuctum goes bilaterally to Edinger-Westphal nuclei

The inputs from the two eyes to the lateral geniculate nucleus do not mix. In what manner are they kept separate?

they go to different layers, contralateral to 1, 4, & 6, ipsilateral to 2, 3, & 5

A thin bar of light made the simple cell fire quickly. Why did a wider bar not do likewise?

Because the wider bar also hit the inhibitory areas in the receptive field of that cell

What technique did Hubel and Wiesel use to get that picture of the visual cortex with stripes?

Autoradiography, or, inject radioactive amino acid into one eye and observe the protein transported transynaptically from ganglion cell across LGN to cortex

As an electrode is advanced obliquely across columns in the visual cortex, what changes?

One could say preferred eye of input or one could say preferred angle of line

Why did Hubel and Wiesel feed their microelectrode amplifier into a loud speaker?

easy to judge firing rate by the sound

Why did a wide line centered on the simple cell's receptive field elicit less of a response than a narrow line?

even though it stimulates the excitatory receptive field, it also stimulates the surrounding inhibitory ones

While an electrode is advanced obliquely through the cat's visual cortex, first the left eye predominates, then the right eye. What else has been changing during that advancement?

preferred angle

When is the critical (sensitive) period for development of binocular connections in the cat visual cortex?

birth to 2 & 1/2 mo

When autoradiography was used to demonstrate ocular dominance columns, how did the film (photographic emulsion) get exposed?

radioactivity (in cortex protein) exposed film in the dark



Return to Syllabus

Return to Stark home page

This page was last updated 3/18/08