Blindsight

Blindsight

1. the ability of people who are perceptually blind to respond to visual stimuli they do not consciously see

   • e.g. pupillary reflex reaction to light
   • type 1: no awareness of stimuli but can still predict better than chance type 2: some awareness but no visual perception

2. term coined by Sanders et al., (1974)

3. caused by damage to primary visual cortex (V1)

   • highlights difference between conscious and unconscious perception and multiple processing streams
     • ventral stream
       • role of vision in perception and visual awareness
         • facilitates recognition and memory
       • ventral stream = damaged in blindsight as V1 is damaged = no visual awareness
       • depends almost entirely on V1 for its inputs
         • U&M (1982): retina > leteral geniculate nucleus > V1 > inferotemporal cortex
     • dorsal stream
       • role of vision in action
       • dorsal stream = intact in blindsight = allows for subconscious perception
       • inputs from V1 + several other subcortical visual structures
         • e.g. U&M (1982): retina > superior colliculus > pulvinar nucleus > posterior parietal cortex
       • BUT not all blindsight requires use of dorsal stream
         • Perenin & Jeannerod (1978): Ps had no intact V1 or any other visual coritcal areas on damaged side of the brain (thalamic tissue ie. LGNd and pulvinar would also have been degenerated)
           • eye movements can be accounted for by the intact retinal pathway to the SC and actions due to cerebral networks
     • Ungerleider and Mishkin (1982): both the dorsal and ventral diverge anatomically from V1
   • also known as striate cortex (usually in animals)
   • damage does not invade either processing stream but would denervate the ventral stream and would leave the dorsal stream (with many assoc structures) intact

4. mapped using perimetry

   • region of blindness = scotoma
     • lesion in the left hemisphere (e.g. left V1) = right visual field deficit & vice versa
   • optic nerve lesion (e.g. between the eyeball and the brain)
   • cortex lesion (ie. V1)
     • hemianopia
   • optic chiasm lesion = complicated.. depends on which fibers are most affected

5. Pupillary response

   • initially dismissed as "purely reflexive"
   • Weiskrantz (1990) found the pupil to be sensitive to variations of stimulus motion, spatial frequency and possibly colour all within a patients "blind" visual field
   • Zihl et al (1980) in the absence of V1, the pupil still shows clear phasic responses to "unseen" light flashes within the scotoma

6. Reaching and grasping

   • Weiskrantz (1986): Ps with cortical blindness in one visual hemifield can use spatial information to guide actions within the blind field
     • Poppel et al. (1973): first reported it when asking Ps to move their eyes towards the light they "could not see"
       • these were inaccurate but still signif related to the location of the light
     • Weiskrantz et al (1974): patient D.B. successfully pointed towards light
       • Weiskrantz et al (1974) and Perry & conwey (1984): attributed the residual visuomotor ability in cortically blind to the ~10% of the optic fibres that terminate in the SP
         • SP = long been implicated in saccadic eye movements (e.g. Sparks and May, 1990)
         • also sends extensive projections to dorsal-stream areas in cortex via pulvinar nucleus of the thalamus
         • Solomon et al (1981): showed SC directly involved through monkey lesions. After V1 lesion it could still reach/grasp but additional colliculus lesion this was lost
     • Perenin and Jeannerod (1975): further support in hemianopic patients
     • Further this skill can be enhanced through training suggets that considerable visual control of the direction and amplitude of eye/arm movements can be developed in cortically blind
       • Zihl (1980): accuracy of saccadic responses was markedly improved
   • Perenin & Rossetti (1993): post a card into an open slot; demonstrated moderately accurate orientation of the card. BUT not when asked to make perceptual judgements
     • could also reach out and grasp objects fairly accurately but approached with a widerhand grip than normal; again, was v. poor at perceptual judgements
     • orientation and size can be processed in cortically blind visual fields only when used to guide a motor action.

7. Detection and discrimination

   • Zihl & von Cramon (1980): after training, hemianopic Ps in reporting when light had come on in their visual field through eyelid response but v.poor at verbally reporting this
     • training could improve visuomotor skills, rather than an impact on general visual sensitivity
     • Marcel (1993) reported similar even in the absence of training. Patient GY most accurate at blinking responses than verbal
     • blindsight Ps might be able to use info derived from visuomotor control systems to guide above-chance performance even on a forced-choice test of detection or discrimination
     • was also accurate in dicriminating between cross/circle/line (70-80% accuracy)
   • Weiskrantz et al (1974): Patient DB described having a "feeling" that stimulus had occurred despite "seeing nothing"; perhaps these feelings are associated with th aciviation of different visuomotor subsystems?
     • This "awareness" and evidence from PET imaging (Barbur et al, 1993) suggests that MT is activated by moving stimuli and thus an awareness it passed back to V1 (as it is in normal human processing)
       • Not all blindsighted patients posess an awareness of motion stimulus due to additional extrastriate damage (e.g. if MT is also damaged there would be little residual visual ability or awareness. if V4 is damaged blindsight would be predicted but not any awareness.) Only where damage spares both ventral and dorsal streams would you expect to see islands of consious vision as in G.Y. (also seen in a "clean" lesion of V1)

8. Colour processing

   • Weiskrantz (1974): DB could guess whether light was green or red
   • Stoerig (1987) 10 cortically blind patients could differentiate between white and coloured light
     • Stoerig & Cowey (1992): evidence for wavelength discrimination
     • colour processing usually associated with parvocellular channels. BUT small number of cells project directly from parvocellular LGN layers to areas V2 and V4. This may subserve colour discrimination
       • Cowey & Stoerig (1992):specialised interneurones survive in the LGNd and receive a direct visual input from the retina and connect to cells which project to areas V2 and V4

9. Motion processing

   • Blythe et al (1987): sensitive to object movement despite blindsight
   • some Ps only show blindsight FOR moving objects
     • Riddoch (1917): early on noted that Ps are better able to "see" an object when in motion in the defective field than if its stationary
   • Barbur et al (1993): under some circumstances, GY does not have blindsight for motion.. he can actually perceive it
   • SC thought to play a role but more importantly instant projections via the pulvinar to dorsal stream (tectopuvlinar pathway to the MT/V5) are required

10. Weiskrantx & Warrington (1973): Patient D.B.

    • damage to RIGHT OCCIPITAL CORTEX
      • hemianopia (half visual field)
      • remarkable recovery after 3 years (perimetry = pac man!)
      • Poppel, Held & Frost (1973): "cant see anything" but eye movements towards light that flashed in the scotoma
        • further, performance improved when asked to point towards light rather than just gauging eye movements
          • support for residual visual functions

11. concept was disputed

    • Campino et al. (1983)
      • could just be a scattering of light in the pupil meaning functioning visual field is picking up light and not scotoma.
        • no perception from damaged visual field area but scattering light produced reduction in vision sensitivity
        • BUT: Weiskrantz (1987) demonstrated that DB was unable to point accurately at stimuli positioned at the blind spot (optic nerve joins retina)
          • light pointed at this spot should have been scattered the same as light pointed towards to scotoma; should also have been able to be perceived if it was ude to faulty-fixation
          • this dissociation is not explained by either the faulty-fixation or scattered-light hypothesis
      • residual functioning cells pick up light due to inaccurate occular fixation
        • Perenin & Jeannerod (1975, 1978): tested patients with severed optic chiasm (ACUTALLY blind) compared to cortically blind (damage in cortex)
          • Could have no residual vision in hemiopia as actual nerve fibres are damaged
            • if its due to light scattering, they should have still been able to perceive stimuli focused on blind field
              • no evidence for blindsight in patients with severed optic chiasm; cortically blind performed well
          • also excluded possibility of faulty-fixation as they excluded trials if P made an ocular deviation of >5degrees
        • MRI imaging shoes no spared regions in scotoma, fMRIs show no residual tissue is activated by the stimulus
        • Cowey & Stoerig (1997): blindsight occurs in monkeys with entire visual cortex surgically removed
          • evidence for two processing pathways and one which bypasses V1
      • still argument for possibly due to inadequate fixation, nonvisual cues or experimenter effects

12. Conclusions

    • blindsight is paradoxical ONLY if vison is considered a unitary process.
      • The mechanisms providing perceptual experience are separate from those providing underlysing visual control of action
      • residual abilities are not due to "unconscious perception" rather due to a collction of visuomotor skills and strategies
    • retinal stimulation does not equal experience of sight, rather it enables access to visuomotor control mechanisms

13. Why is it unconscious??

    • mediated by two pathways
      • retino-tectal pathway (superior colliculus > pulvinar): mediates residual motion & location detection and direct route for saccadic eye movements
        • This primitive SC circuitry is inadequate for "awareness" also needs cortex
      • geniculo-striate pathway (LGNd > V2, V4): mediates residual form and colour discrimination
      • the weak visual signal produced by these pathways is below "consciousness" threshold (Kalat, 2009)
    • Milner & Goodale (1995): damage to V1 responsible for unconsciousness. tectopulvinar route ( SC>pulv>extrastriate>dorsal strea) still receives input and so can still control eye/arm movements
      • BUT: is blindsight just a collection of visuomotor responses elicited by the SC and dorsal stream?
        • NO: Perenin & Rosetti (1993): evidence for action-related processing. patients reach out and grasp object relatively accurately in blind field
          • BUT: blindsight in still demonstrated in Ps with only one hemisphere and thus no striate, no LGN or pulvinar (children who have surgery)
            • Conwey & Stoerig (1987,1992): colour discrimination after V1 lesions. This disputes milner and goodales explanation as visuomotor response (ie.SC & dorsal processing = SC has no orientation specificity or colour processing)
              • could be explained by v.small number of cells that input from parvo layes of LGN directly to V2 & V4 (ventral pathway) and skip past V1/damaged area. unclear what brain function this pathway would have in regular processing

14. Inducing Blindsight (through productive TMS)

    • blindsighted Ps: stimulate V5 = report no induction of conscious visual sensations (Patieng GY). V1 is essential for conscious perception
    • peripherally blind: stimulate V1 or V5 = induction as much as normally sighted (patient PS). Because V1 is still in tact, stimulating it produces sensations of light
    • Pascual-Leone & Walsh (2001): TMS of V1 abolishes conscious perception of V5-TMS when virtual lesion occurs after V5-TMS
      • e.g. although when V5 is stimulated = perception, if lesion is then induced (TMS) to V1, this conscious perception disappears (supports evidence that V1 is essential for conscious perception)

Images and files in the map