*Light hits the retina and activates/deactivates retinal ganglion cells according to their distribution, sending signals along the optic nerve
*Most axons of the optic nerves terminate in the lateral geniculate nucleus and the superior colliculus, in the thalamus; superior colliculus helps coordinate rapid eye movements toward a target
*Optic radiations from these structures arrive at the striate cortices (i.e. V1, or the primary visual cortices), where "most" visual information eventually reaches
*The distribution of neurons in the V1 represents a topographic map corresponding to the distribution of retinal cells, which cluster most heavily around the fovea (i.e. retinoptic mapping)
*The V1 works in
parallel with other parts of the brain (i.e. extrastriate cortex) to process form, color, location, movement, etc.
*The extrastriate regions also project a map of the retina
*The human V1 is located in a different anatomic position than the monkey V1.
*Gaps in the visual field (perceptually) are called
scotoma and can be predicted based on information pertaining to which part of the visual pathway is damaged; these gaps are not perceived as just dark patches, but rather places where nothing can be perceived at all, though there can be subconscious "blindsight" (c.f. "hemispatial neglect", which is neglect of the side opposite to an injured cerebral hemisphere).
*Sensory cells have receptive fields that constitute a "stimulus region" and the places/features that excite or inhibit the cell
*On-center bipolar cells in the retina are excited/deinhibited by light in the center of their receptive fields; for off-center bipolar cells, turning off light hitting the center of their receptive fields activates them
*Bipolar cells behave oppositely to their adjacent neighbors, which are also arranged concentrically, and are connected to further ganglion cells of the same type that when activated report a binary "light or dark" to vision-centers in the brain through the optic nerves
*The 6 layers of the LGN also have concentric receptive fields; the large-cell layers are not involved in color discrimination, while the small-cell layers are
*The V1 does not respond to the same stimuli that activate the LGN cells that project into it; it needs more complex stimuli
*V1 cells differentiate (in addition to the retinoptic mapping) by responding best to lines in particular orientations and positions
*Some require movement of stimulus for optimal response; the motion must respond to the same criteria of orientation and position (i.e. direction)
*The hierarchical model had it that simple cortical cells receive input from a number of LGN cells, and generate information based on the input and their criteria for activation; complex cortical cells receive input from a number of simple cortical cells and construct more-complex events from their input
*However, the V1 is not hierarchical enough to account for all human object-recognition/construction in terms of the construction of complexity from increasingly-complex inputs, as there are not enough neurons or levels of neurons
*The
spatial-frequency filter model replaces the hierarchical model for patter-analysis/object-construction-recognition; something about Fourier analysis/linear-systems theory and application to both auditory and visual processing in terms of sine waves
*Channels in the visual system for different spatial frequencies
*This model would explain the ambiguity of the Mona Lisa expression (happy or sad?) thus:
The foveal vision reports the high-frequency parts of the image. The Mona Lisa with low-frequency components filtered out looks sad (Google this). The peripheral vision can only detect low-frequency parts. The Mona Lisa with high-frequencies filtered out looks happy. Spatial-perceptual ambiguity creates artistic effect.
*V1 neurons are also important to the perception of mental images; when you see an imaginary image, there is retinoptic activation of select parts of the V1 (e.g. small imaginary things show up in the imaginary foveal vision).
*Repeated magnetic stimulation to V1 impairs formation of mental imagery
*From V1, information travels to extrastriate (i.e. non-primary) parts of the visual cortex, including V2-5, which also receive some input from the LGN (though I'm not clear on whether this is exclusive input that doesn't make it directly to the V1)
*There may in fact be dozens of distinct cortical areas (in addition to the V1) that deal directly with visual information/function, more areas than the number of basic attributes; only a few are being described in detail here
*30% of the surface of the human cerebral cortex is largely or entirely devoted to the analysis of visual information
*V2 is adjacent to V1 and responds to optical illusion more readily than V1
*V4 receives axons from V2 and responds strongly to sinusoidal, concentric, and radial spatial-frequencies, as well as to differences in light wavelength
*If V1, V2 deal with spatial frequencies, and the inferior temporal area deals with pattern and form, then the V4 may be intermediate in function between them
*V5/Medial-temporal area is specialized for motion-perception
*The inferior temporal area responds to specific complex forms, especially learned forms
*Is not activated so much by faces; a part of the prefrontal cortex responds to faces and nothing else, and receives connections from the inferior temporal gyrus as well as the superior temporal sulcus
*skip a few sections I deem superfluous at this point*
*Most of the cortical visual areas are organized into two streams that originate in V1 (though there are many cross-connections between them)
*The ventral stream identifies objects (
what-stream) and the dorsal stream locates them (
where-stream)
*In the ventral stream, from the occipital to the frontal lobes, information about faces becomes "more specific"; representation of individual faces is more anterior, while general facial structure (and gender identification) is more posterior
*Ataxia - damage to dorsal stream causes difficulty in moving towards objects and/or reaching to grasp them, but leaves identification intact; the opposite disorder has also been observed
*The anterior dorsal stream merges with the primary motor cortex and includes neurons with dual visual-motor functions (known as mirror-neurons)
*Mirror neurons in the premotor cortex may mediate the understanding of actions performed by other creatures, transforming visual information into knowledge that can be applied by the learning organism
*Differences between monkey and human mirror-neurons include:
Human mirror-neurons are activated by "meaningless movements" and monkey-neurons aren't. Human mirror-neurons break down movements into components while monkey-neurons only code the total movement.
Bookmarks