Neurophysiology LOs Flashcards
What are the problems associated with using templates as feature detectors
Corners can be detected by light falling on the receptors comprising a feature template
Problem 1) a uniform bright field, devoid of corners will also set off receptors surpassing threshold and making explicit a corner where there is none (when just positive receptors)
Solution: detectors must be more sophisticated, make some connections inhibitory so when light falls on it creates negative response
Problem 2) Is there just one upper-right corner detector in the eye? What if the image of an upper right corner doesn’t fall on this single detector?
Solution: receptive fields on retina are overlapping, any given receptor is not connected to a single feature detector, are connected to many
Problem 3) How does our corner detector handle edges, Yuck edges can cause detector to exceed threshold
Solution: Increase the threshold
Problem 4) If the threshold is increased, dim corners will not be detected.
OVERALL: No satisfactory trade-off exists. Template based systems are intrinsically ambiguous, detecting complex features with templates if flawed
Describe Structures comprising the different pathways of primate visual system, from retina to LGN to cortex
M-Path: photoreceptors (rods/cones) -> diffuse bipolar cells -> parasol ganglion cells (M-Cells)
P-Path: photoreceptors (cones) -> midget bipolar cells -> midget ganglion cells (P-Cells)
Ganglion cells project to LGN which is in the thalamus to 2 major subidvisions (Magnocellular and Parvocellular also Koniocellular)
Magnocellular: 2 ventral large cell layers, input from large Retinal Ganglion M-Cells, found in Mammals.
Not wavelength sensitive, 2-3x larger receptive fields than parvo better visual acuity, respond faster, faster rate of decay, better at handling motion, more sensitive to low contrast (1-15% difference in brightness)
Parvocellular: 4 dorsal small-cell layers, input from small P-Cells found in primates
90% Selectively wavelength sensitive, better at resolving spatial detail, sensitive to low and high contrast (respond from 1%)
Koniocellular: smallest LGN cells located between parvo and magno layers, input from bistratified RGCs, only in primates
10-20% of LGN connections to LGN are from retina, 50% from higher cortical areas (feedback mechanisms) rest from local inhibitory and brainstem inputs
Retinal Ganglion Cell axons also go to superior colliculus (visual reflexes), pretectum (pupillary reflexes), accessory optic system (perception of self motion/gaze stabilization), suprachiasmatic nucleus (circadian rhythms)
Magno LGN -> V1 layer 4Ca -> 4B -> thick stripes (v2), V3 and MT (V5)
4B cells are orientation selective, selective for motion in a particular direction, have binocular sensitivity, NO colour sensitivity, involved in motion/spatial analysis
Parvo LGN -> V1 layer 4cB -> interblobs (layer 2/3) -> pale stripes (V2) -> V4, involved in processing form
Parvo LGN -> V1 layer 4CB -> blobs (layers 2/3) -> thin stripes (V2) -> V4, involved in processing colour
Konio LGN -> blobs (other regions including MT)
What is the nature of the functional division in the visual system
Evidence for subdivision of visual perception: why? -> segregation into subsystems allows for greater specialization and more complex processing of the visual scene, redundancy is adaptive
Motion: found by using equiluminant stimuli; same intensity (amount of light) diff wavelengths, motion perception of equiluminant red and green sinewave stripes impaired. Slowly rotating disk with stripes of 16 or 32 cycles/degree often appeared to be stationary, motion perception impaired at high spatial frequencies
Depth: pictorial depth cues rendered with equiluminant colours: depth perception eliminated or diminished, all elements of image are visible, implies low-level processes are responsible
Figure-Ground Discrimination: linking by collinearity (illusory contours) breaks down with equiluminant stimuli, without equiluminant see triangle, with equiluminant see 3 pacman meeting at intersection, may be due to inactivation of depth mechanism
Magno and parvo system function complementarily!
Magno: motion detection, temporal analysis, depth perception, more primitive
Parvo: form analysis, spatial analysis, colour vision, recently evolved
Describe organization of visual cortex and beyond
Blobs: pillar like cortical sections, found in V1 layers 2 and 3, revealed by cytochrome oxidase staining of mitochondria, input from parvo and konio, no orientation selectivity, have monocular sensitivity, wavelength/brightness selective, most doubly colour opponent
Interblobs: areas between blobs in V1 layers 2 and 3, input from parvo, cells are orientation selective, have small receptive fields, have binocular sensitivity, not wavelength selective, or to direction of movement
Simple cell: in V1 layer 4, responds best to: bar/line, in a particular location on retina, having specific orientation, constructed by converging centre-surround cells
Complex Cell: in V1 layers 2,3,5,6 V2. Respond best to: bar/line, in particular location on retina, have specific orientation, and moving in a certain direction
Hyper Complex / End-stopped cell: V2 and V3, respond best to: bar, corner or angle having certain length/width, in particular location on retina, having specific orientation and moving in a specific direction
When electrode inserted perpendicular to surface of cortex cells respond to stimuli having the same property: location column (same retinal location), ocular dominance column (stimuli by one eye only), orientation column (same orientation, adjacent columns are selected by about 10 degrees)
Hypercolumn: 2 mm x 2 mm region containing single location column, left and right ocular dominance columns and a set of orientation columns from 0 - 180 degrees.
How do pinwheels Relate to the packing problem
Packing problem: how are 3 parameters of info (X position, Y position and orientation) mapped onto a 2D representational space
Cortical depth is not used to encode diff orientations as cells within a column all respond to the same orientation
Pinwheel patterns of orientation columns revealed within given ocular dominance column, found by phtoographing cortex with diff stimuli and superimposing images
Iso-orientation contours (lines that have same orientation preference) radiate from singularities (lying in centre of ocular dominance bands)
Retina does not map to cortex in isomorphic point-to-point fashion
What evidence is there for separate streams of processing in extrastriate visual areas?
Dorsal stream/Parietal Pathway (where/how): receives mostly magno input, neurons in MT/V5 appear to be involved in analyzing speed and direction of stimulus motion and stereoscopic depth analysis, projects to posterior parietal association cortex, stream seems to be processing aspects of location of visual stimuli, neurons in MST involved in visual navigation, directing eye movements, motion perception.
Implicated in akinetopsia (inability to perceive motion of objects)
Ventral stream/inferotemporal pathway (what): receives mix of magno and parvo, V4 involved in colour processing, V4 cells also orientation selective, suggesting form analysis, stream for processing shape and colour. IT involved in processing complex spatial arrangements
V4 implicated in cerebral achromatopsia (deficit in colour vision without loss of object perception)
IT implicated in prosopagnosia (loss of ability to recognize faces with otherwise normal vision)
Balint’s Syndome: bilateral damage to superior posterior parietal lobes. Exhibited optic ataxia (inability to reach/grasp objects), optic apraxia (inability to guide eye movements or change visual fixation), simultagnosia (inability to perceive more than one aspect of visual stimulus and integrate details into coherent whole), could recognize objects but could not tell where located or reach for them. Intact what but damaged where
Visual Form Agnosia: inability to extract global structure, despite intact low-level sensory processing. Can’t recognize/discriminate/copy complex visual forms like shapes, can grasp objects they can’t identify, intact where system but damaged what system
See Double dissociation, Below shows DD within intact brains, Above shows double dissociation of what vs where
Also see an illusion of circle surrounded by small circles and circle surrounded by big circles. First one seemes bigger. Visually size of centre disk is affected by surrounding discs. When asked to pick up centre disc grip apperture was perfectly calibrated to exact size od disk. Observers treated discs that were perpetually different as physically identical showing visual perception and visually guided action are separate
Dorsal: specialized to interact with objects (same every time)
Ventral: specialized to identify objects from various POVs (shape constancy)
