Visual acuity: oh say, can you see?

  • Acuity: the smallest spatial detail that can be resolved o Vision scientists: the smallest visual angle of a cycle of grating that we can perceive o Eye doctors: 20/20 vision is perfect (distance at which the person can just identify the letters/distance at which a normal person can just identify the letters)
  • Cycle: one repetition of a black and a white stripe
  • Visual angle: the angle that would be formed by lines going from the top and bottom (or left and right) of a cycle on the page, through the center of the lens, and onto the retina o Arctan (size of the cycle/viewing distance)
  • The spacing of photoreceptors on the retina determines the limit of spatial vision, the finest high-contrast detail that can be resolved
  • Aliasing: misperception of a grating due to undersampling; if the light intensity of gratings varies smoothly and continuously across each cycle, the visual system samples it as discretely through the array of receptors at the back of the retina. If the receptors are spaced in such a way that the whitest and blackest parts of the grating fall on separate cones, we would be able to make out the rating. But if the entire cycle falls on a single cone, we perceive a grey field instead of stripes.

 

Acuity for low-contrast stripes

  • Spatial frequency: the number of grating cycles in a given unit of space o Cycles per degree: the number of pairs of dark and bright bars per degree of visual angle
    • Schade: let participants adjust the contrast of the gratings of a specific spatial frequency until they could just be detected
  • Contrast sensitivity function (CSF): a function describing how the sensitivity contrast (defined as the reciprocal of the contrast threshold: the smallest amount of contrast required to detect a pattern) depends on the spatial frequency of the stimulus; shaped like an inverted U o Reciprocal of the contrast threshold: for a 1-cycle/degree grating to be just distinguishable from uniform grey, the dark stripes must be about 1% darker than the bright stripes, the reciprocal of this threshold is 1/0.01 = 100

Retinal ganglion cells and stripes

  • Retinal ganglion cells are tunes to a specific spatial frequency that matches its perceptivefield size (because then a bright bar fills the center and dark bars fill the surround), and it responds less to higher (where the ON-center ganglion cell will respond weakly because both dark and bright stripes will fall within the receptive-field center, washing out the response) and lower (where the ON-center ganglion cell will respond weakly because part of the fat bright bar of the grating lands in the inhibitory surround, damping the cell’s response) spatial frequencies
  • The phase of the grating, its position within the receptive field, also influences the response of a retinal ganglion cell. When the grating has a bright bar filling the receptive-field center and dark bars filling the surround, the ON-center ganglion cell will fire vigorously. However if the grating phase is shifted 90 degrees, half the receptive-field center will be filled by a bright bar and the other half will be filled by a dark bar, and similarly for its surround. Another 90 degree shift will cause a dark bar to fall on the receptive-field center, and bright bars on the surround, which will lead to a negative response of the ON-center ganglion cell.

The lateral geniculate nucleus

  • Magnocellular layers: the bottom two o Respond to large, fast-moving objects
  • Parvocellular layers: the upper four o Processes details of stationary objects
  • Koniocellular layers: between magno- and parvocellular layers
  • Left LGN: receives projections from the left side of the retina in both eyes
  • Right LGN: receives projections from the right side of both retinas
  • Input from the contralateral eye is received by layers 1,4 and 6 of the LGN
  • Input from the ipsilateral eye is received by layers 2,3 and 5 of the LGN

Striate cortex

  • Striate cortex: receiving area for the input of the LGN (primary visual cortex V1)
  • Each part of the visual field has a specific region in the striate cortex, with more important parts having larger regions (near the fovea larger than far sight of the periphery) →cortical magnification figure 3.15 page 66

Some perceptual consequences of cortical magnification

  • With eccentricity (increasing distance from the fovea) visual acuity declines
  • Visual crowding, the deleterious effects of clutter on peripheral object recognition, is the main obstacle to reading or object recognition in the periphery rather than the fall off of visual acuity o Crowding is a process that simplifies the appearance among adjacent objects, at the expense of the ability to pick out individual objects.

Receptive fields in striate cortex

  • Receptive fields of striate cortex neurons are not circular, as they are in the retina and the LGN, rather they are elongated. As a result they respond much more vigorously to bars, lines, edges and gratings than to round spots of light.

Orientation selectivity

  • Orientation tuning: the tendency of neurons in the striate cortex to respond to certain orientations and less to others → the cell is tuned to detect lines in a specific orientation
  • There are more cells responsive to horizontal and vertical orientations than to obliques → humans have a somewhat lower visual acuity and contrast sensitivity for oblique targets than for horizontal and vertical targets
  • Striate cortex cells respond to lines, because the input they receive from the LGN cells are all in a row

Other receptive-field properties

  • Ocular dominance: striate cortex neurons respond to stimuli that are in a specific directions, regardless whether it is presented to the left eye or the right. However they do tend to have a preference, responding somewhat more rapidly when a stimulus is presented to one eye than when it is presented to the other.
  • Striate cortical cells respond well to gratings, and they respond best to gratings that have just the right spatial frequency to fill the receptive-field center → each cortical cell is tuned to a particular spatial frequency, which corresponds to a particular line width.

Simple and complex cells

  • Simple cell: a cortical neuron with clearly defined excitatory and inhibitory regions o Edge detector: light on one side of its receptive field, darkness on the other o Stripe detector: responds to a line of light that has a particular width, surrounded on both sides by darkness
  • Complex cell: a neuron whose receptive-field characteristics cannot be easily predicted by mapping with spots of light
  • A complex cell will respond to light in a particular orientation and spatial frequency and shows an ocular preference. However where a simple cell might respond only if a stripe is presented in the center of its receptive-field, a complex cell will respond regardless of where the stripe is presented as long as it is somewhere within the cell’s receptive-field

Further complications

  • End stopping: the process by which a cell in the cortex first increases its firing rate as the bar length increases to fill up its receptive-field, and then decreases its firing rate as the bar is lengthened further o Plays an essential role in our ability to detect luminance boundaries and discontinuities
  • The size of a particular cell’s receptive-field appears to vary with target contrast o The cell might respond to a smaller portion of the visual field when the grating stimulus has a high contrast as opposed to when the difference between light and dark bars is more subtle
  • Neurons can be influenced by stimuli that fall outside the classic receptive-field, via short- or long-range lateral connections and/or via feedback from neurons in other layers

Columns and hypercolumns

  • Neurons with similar orientation preferences are arranged in columns that extend vertically through the cortex
  • Neurons that share the same eye preference also have a columnar arrangement
  • Hypercolumn: a 1-mm block of striate cortex containing two sets of columns, each covering every possible orientation (0-180 degrees), with one set preferring input from the left eye and one set preferring input from the right eye. However not all hypercolumns see the world at the same level of detail due to cortical magnification.
  • Cytochrome oxidase (CO): an enzyme used to reveal the regular array of the CO blobs, which are spaced about 0.5mm apart in the primary visual cortex. CO blob columns have been implicated in processing colour, with the interblob regions processing motion and spatial structure.
  • The striate cortex is concerned with analysing the orientation, size, shape, speed and direction of motion of objects in the world, and it does so by using modular groups of neurons –hypercolumns- each of which receives input from and processes a small piece of the visual world.

Selective adaptation: the psychologist’s electrode

  • Selective adaption: a reduction in response caused by prior or continuing stimulation, due to fatigue in the right degree cells your perception shifts the true vertical line a few degrees to the side because those cells have a higher firing rate now o Can provide insights in the properties of cortical neurons
  • Tilt aftereffect: the perceptual illusion of tilt, produced by adaptation to a pattern of a given orientation
  • Figure 3.30 page 78

The site of selective adaptation effects

  • Interocular transfer: when you do the adaptation experiment with one eye, and look back at the original picture with the other eye, the tilt aftereffect still occurs.
    • The transfer of adaptation from one eye to the other implies that selective adaptation occurs in the cortical neurons, because the info from both eyes is kept completely separate in the retinas and the two LGNs: no single neuron receives info from both eyes until the striate cortex

Spatial frequency-tuned pattern analysers in human vision

  • Spatial-frequency channel: a pattern analyser, implemented by an ensemble of cortical neurons, in which each set of neurons is tuned to a limited range of spatial frequencies o We have fewer neurons tuned to low spatial frequencies
  • Multiple-spatial-frequency model of vision implies that spatial frequencies that stimulate different pattern analysers will be detected independently, even if the different frequencies are combined in the same image o The contrast sensitivity for a compound pattern that is made out of two other compound patterns is almost the same as the contrast sensitivity for detecting the individual components of the pattern separately → the visual system filters the image into spatially localized receptive fields that have a limited range of spatial frequencies
  • Different spatial frequencies emphasize different types of info o Low frequency: broad outlines of an image o High frequency: fine details of an image

The development of spatial vision

  • Preferential looking is used to determine what infants can see o The infant will prefer the more complex scene, if the infant doesn’t prefer stripes over uniform grey this implicates that it cannot see the stripes

Development of the contrast sensitivity function

  • Contrast sensitivity at low spatial frequencies is fully developed as early as 9 weeks, whereas sensitivity at high spatial frequencies continues to develop dramatically o From birth to beyond 4 years of age, cone density increases in the central region because of both the migration of receptors and decreases in their dimensions, which results in finer cone sampling

The girl who almost couldn’t see stripes

Congenital cataracts: one eye does not receive normal stimulation in the critical period of early visual development, therefor the neurons that should be destined to respond to that eye do not become properly connected, making visual acuity impossible.