Brain, Vol. 124, No. 1, 30-46,
January 2001
© 2001 Oxford University Press
Motion discrimination in cortically blind patients
Department of Experimental Psychology, University of Oxford, Oxford, UK
Correspondence to:
Dr Paul Azzopardi, Department of Experimental Psychology, University of Oxford, South Parks Road, Oxford OX1 3UD, UK E-mail: paul.azzopardi{at}psy.ox.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
Some patients with brain damage affecting the striate cortex, though clinically blind in their field defects, can still discriminate visual stimuli when forced choice procedures are used. Such patients seem particularly sensitive to moving stimuli in their scotomata, though there are conflicting reports as to whether they can discriminate the direction of motion. We tested three patients with areas of cortical blindness for their ability to detect and discriminate the direction of motion of a variety of first-order motion stimuli, namely bars, gratings, plaids and random dot kinematograms depicting translation and motion in depth, during forced choice tasks. The patients could detect the presence of movement in any kind of stimulus, and could discriminate the direction of single bars, but none could discriminate the direction of motion of the more complex stimuli (gratings, plaids and random dot kinematograms) or discriminate between 0 and 100% coherent random dot kinematograms at any speed tested (from 4 to 64°/s). Similar results were obtained from one of the patients who was additionally tested with second-order versions of the translated bar and random dot kinematograms, eliminating light scatter as an explanation. Overall, the results suggest that motion processing in the scotoma is severely impaired, and that the puzzling discrepancies between previous studies can be accounted for by the type of stimulus used. The motion discrimination impairment caused by brain damage affecting the primary visual cortex is inconsistent with the proposed existence of a subcortical pathway to extrastriate cortical motion areas (such as areas MT and MST) which bypasses the striate cortex and is specialized for analysing `fast' motion.
blindsight; visual motion perception; area V1; area V5/MT; superior colliculus
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
The striate cortex (area V1) contains a topographic map of the visual field, such that locally damaging or disconnecting it causes blindness in the corresponding part of the visual field (Holmes, 1945
). Some patients with a damaged striate cortex, though clinically blind as defined by perimetry, can detect and even discriminate between stimuli presented in their field defects when forced choice procedures are used, even though they deny seeing them (Pöppel et al., 1973; Weiskrantz et al., 1974). This phenomenon is known as blindsight. The range of visual capacities spared following striate cortex lesions includes the ability to locate and discriminate high-contrast targets by eye movements, by pointing and by verbal report (for reviews, see Weiskrantz, 1990; Stoerig and Cowey, 1997). Evidence from monkeys with striate cortex lesions and from brain-damaged patients suggests that these capacities are mediated by neural projections from the retina to the extrastriate visual cortex that bypass the predominant route from the retina to the striate cortex via the lateral geniculate nucleus, and which involve the superior colliculus and the lateral geniculate and pulvinar nuclei of the thalamus (Mohler and Wurtz, 1977; Rodman et al., 1989, 1990; Gross, 1991; Cowey and Stoerig, 1991; Bullier et al., 1994; King et al., 1996; Azzopardi et al., 1996).
Moving stimuli are usually more readily detected in the field defects of cortically blind patients than static ones. In fact, moving targets may be so salient that, in carefully controlled conditions, it can be difficult to discern whether the patients are actually `blind' to them (Riddoch, 1917; Holmes, 1918; Weiskrantz, 1990; Weiskrantz et al., 1995; Azzopardi and Cowey, 1998; Zeki and ffytche, 1998). Whether this reflects sensitivity to motion is open to question, for two reasons. First, moving stimuli have not always been equated for detectability with stationary stimuli in normal vision, i.e. moving targets could be detectable in the field defect because they are more easily detected than static targets in normal vision. Secondly, motion may be confounded with position and temporal frequency, both of which can be discriminated in the field defect independently of motion (Pöppel et al., 1973; Barbur et al., 1994). Several studies report that cortically blind patients can discriminate explicitly the direction of a variety of moving stimuli, including single spots (Blythe et al., 1986, 1987; Weiskrantz et al., 1995; King et al., 1996), bars (Barbur et al., 1993), gratings and plaids (Perenin, 1991; Benson et al., 1998; Morland et al., 1999), and random dot kinematograms depicting translation (Perenin, 1991; Benson et al., 1998; Zeki and ffytche, 1998) or motion in depth (Mestre et al., 1992). But others have found them unable to discriminate the direction of motion of gratings and random dot kinematograms depicting translation, relative motion and motion in depth (King et al., 1996; Barton and Sharpe, 1997), and monkeys with striate cortex lesions cannot discriminate the direction of moving gratings presented in the scotoma (Weiskrantz, 1963).
The idea that motion processing is preserved in the scotoma is bolstered by studies of the properties of neurones in extrastriate cortical visual areas in monkeys with striate cortex lesions, especially area V5/MT, which is specialized in its sensitivity to motion direction (Dubner and Zeki, 1971; Albright, 1984). Recording from this area, Rodman and colleagues found that the visual responses of single neurones with receptive fields in the scotoma were weaker and more variable than normal after surgical or reversible lesions, yet about half of the neurones sampled still retained the ability to discriminate the direction of motion of bars swept through their receptive fields (Rodman et al., 1989). This finding was confirmed, and extended to include area V3a (another visual area specialized for motion), by Girard and colleagues (Girard et al., 1991, 1992). These were unexpected findings, given that few neurones in the superior colliculus, upon which residual motion sensitivity in area MT depends (Rodman et al., 1990), are directionally selective (Goldberg and Wurtz, 1972), and Gross (1991) therefore suggested that neurones in area MT must be able to compute directionality locally on the basis of non-directional information supplied by neurones in the superior colliculus.
Given the evidence from single-unit recordings, it would be tempting to ignore those studies which did not find cortically blind patients able to discriminate direction of motion in the scotoma. A recent development, however, is the finding that neurones in cortical areas MT and MST of monkeys with long-standing striate cortical lesions, though sensitive to the presence and direction of motion of isolated moving bars in the scotoma, are nevertheless insensitive to the direction of motion in 100% coherent random dot kinematograms (Azzopardi et al., 1998; Fallah et al., 1998). This suggests that motion processing in the scotoma could be impaired, and that the choice of stimulus could be critical in determining whether or not patients with blindsight are capable of discriminating direction of motion. The purpose of the present work was to investigate this by testing for direction discrimination using both isolated bars and random dot kinematograms in the same patients, which has never been done before, and with the same stimulus parameters as used by Azzopardi and colleagues (Azzopardi et al., 1998), for direct comparability with the monkey brain. The results show that cortically blind patients, like MT neurones in cortically blind monkeys, are severely and selectively impaired in their ability to discriminate motion depicted by random dot kinematograms and gratings. This could have important implications for understanding the functional architecture of motion perception in the brain.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
Three subjects with occipital lobe brain damage were tested for their ability to detect and discriminate the direction of motion of single bars, random dot kinematograms, gratings and plaids, defined either by luminance contrast (first-order motion) or by static or dynamic texture contrast with no associated luminance cues (second-order motion) (Chubb and Sperling, 1988
; Cavanagh and Mather, 1989), using forced choice tests. The subjects' consent to take part in the experiments described was obtained in accordance with the Declaration of Helsinki and the experiments were approved by the Ethics Committee of the Department of Experimental Psychology, University of Oxford. |
Subjects |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
The subjects, G.M., G.Y. and S.P., had dense homonymous field defects caused by unilateral damage to the contralateral medial occipital lobe.
Subject G.Y.
G.Y. was a 42-year-old man with a unilateral lesion in the left medial occipital cortex caused by a traffic accident when he was aged 8 years. The damage caused a right, homonymous haemianopia, with macular sparing extending 3.5° into the otherwise blind hemifield (Barbur et al., 1980). His residual visual capacities included the ability to detect, localize and discriminate transient stimuli presented in his field defect (Barbur et al., 1980, 1994; Blythe et al., 1986, 1987; Weiskrantz et al., 1991; King et al., 1996), many of which he is unaware of (Weiskrantz et al., 1995). His sensitivity is not necessarily mediated by extra-ocularly scattered light (King et al., 1996; Azzopardi and Cowey, 1997) or by spared islands of cortex in his field defect (Kentridge et al., 1997). His lesion is illustrated in Fig. 1. The only remaining striate cortex in his left hemisphere is at the occipital pole, which corresponds to his macular sparing. All the tests described were carried out with stimuli 20° in diameter, centred at an eccentricity of 20° to the right of and 12° above the point of fixation, and viewed monocularly.
|
Subject G.M.
Subject G.M. suffered a mild stroke that destroyed his left ventral striate cortex (Fig. 1
) when he was aged 69 years, causing a contralateral right superior quadrantanopia that was stable over repeated sessions of static and dynamic perimetry. Testing took place 3–4 years after the stroke. Unless otherwise specified, tests were carried out with stimuli 20° in diameter, centred 30° to the left of and 15° above the point of fixation, and viewed binocularly.
Subject S.P.
Subject S.P. (Fig. 1) suffered a stroke when aged 23 years. Her left upper quadrantanopia was only discovered 20 years later and her residual vision was first examined a year after that. The lesion extended more rostrally than in G.Y. and G.M. and included areas of ventral extrastriate cortex that were spared in G.M. and G.Y. The stimuli were centred 30° to the left and 25° above the point of fixation, and viewed binocularly.
Stimuli
Two systems were used to generate and present the visual stimuli. Some stimuli (first-order bars, random dot kinematograms and gratings, and second-order bars and random dot kinematograms with dynamic contrast) were generated using custom software running on a Toshiba Satellite 100CS laptop computer with a 75 MHz Pentium processor and a standard SVGA video card with 6-bit grey level resolution. They were displayed with gamma correction on a 15 inch VDU [0.27 mm pitch, 1024 x 760 pixels in a viewing area of 255 x 205 mm (Chuntex Electronics, Taipei, Taiwan, Model 1565), calibrated with a Minolta LC1500 photometer] at a non-interlaced frame rate of 80 Hz. The remaining stimuli (first-order plaids and gratings, and second-order static contrast bars and random dot kinematograms) were generated using custom software running on a Dan IBM PC-compatible computer with a 200 MHz Intel Pentium MMX processor and a VSG 2/4 visual stimulus generator (Cambridge Research Systems, Rochester, UK). They were displayed, with gamma correction on an Eizo T660 20-inch monitor (0.31 mm pitch, 640 x 480 pixels in a viewing area of 290 x 220 mm, calibrated with OptiCal system from Cambridge Research Systems), with a non-interlaced frame rate of 80 Hz.
All moving stimuli were presented at speeds of 4, 20 32 and 64°/s. Apart from slight differences in contrast and the use of two additional speeds (32 and 64°/s), the parameters of the first-order stimuli were identical to those used by Azzopardi and colleagues in their study of the properties of neurones in area MT after removal of parts of the striate cortex in macaque monkeys (Azzopardi et al., 1998).
First-order stimuli
Translation.
Isolated bars were used to test for motion detection (moving bar versus blank screen) and motion direction discrimination (up versus down). The stimulus (Fig. 2) was a horizontal white bar, 20.0° x 1.0° in size (contrast 0.75 and/or 0.99), presented against a dark background of 0.03–0.3 cd/m2, which moved up or down at speeds of 4, 20, 32 and 64°/s. It was implemented by panning the SVGA image by an appropriate amount each frame at the screen refresh rate of 80 Hz in synchrony with the vertical retrace. The monitor was laid on its side to achieve vertical motion. The bar was not visible in the intertrial interval, but appeared at the beginning of one or other interval behind a 20° x 20° square window cut out of a sheet of black card placed directly in front of the VDU. The stimulus was presented for as long as it took to traverse the window (5, 1, 0.625 and 0.313 s for the four speeds, respectively), and therefore when the bar was paired with blank presentations (during tests of detection) care was to taken to ensure that the duration of the blank was the same as that of the moving stimulus.
|
Random dot kinematograms that depicted translation, in which the motion of the component dots was 100% coherent and the dots had an `infinite' lifetime, were used to test for motion detection (moving versus static stimuli) and motion direction discrimination (up versus down), as shown in Fig. 3
. They consisted of an image of 0.5° diameter dots, painted white on a dark background (luminance 0.03–0.3 cd/m2, contrast range 0.75–1.00) at random positions on the screen and at an average density of 0.414 dots/deg2. The image was viewed through a 20° diameter circular window cut out of a sheet of black card placed directly in front of the VDU. The SVGA image was animated by panning the image at the screen refresh rate of 80 Hz (as described above for bars), which wrapped around the edges of the screen without any visible discontinuity. Each presentation lasted 1000 ms. The image was not erased or redrawn or blanked between trials in order to avoid confounding motion onset with luminance onset, though control tests showed that blanking the image between trials made no difference to the subjects' ability to discriminate the stimuli. The image was redrawn every 10–15 trials as a precaution against the possibility of random, but nevertheless potentially salient, clusters of dots being used as a cue to solving the task (though there is no evidence that this ever occurred).
|
Random dot kinematograms in which the component dots had a limited lifetime of 220 ms were used to test for motion detection (moving versus static stimuli) and for the ability to detect and discriminate coherent motion from random noise (100% versus 0% dot coherence). The stimuli were implemented on the SVGA system by means of double buffering, after Newsome and Pare (Newsome and Pare, 1988
). Initially, each dot in the display was drawn at a random position on the screen and assigned a direction (e.g. up or random) and an age (a random value between 0 and 220 ms). In each subsequent frame the position of each dot was updated according to its assigned speed and direction and its age was incremented by the frame duration. If a dot's age exceeded its limit (220 ms), it was assigned a new location at random, a new direction, and its age was reset to zero. In the static displays, the individual dots did not move, but they were randomly relocated at the end of their lifetime creating a twinkling effect with no overall motion direction. In our implementation, dot positions were updated synchronously at a rate of 22.4 Hz.
Gratings (Fig. 5) were used to test for motion detection (moving versus static grating) and direction discrimination (up versus down). They consisted of an image of an 0.5 c.p.d. (cycles per degree) square-wave grating with a mean luminance of 10 cd/m2 and a contrast of 1.0, viewed through a 20° diameter window in a sheet of black card placed directly against the VDU. The SVGA image was animated by panning the image (with wrap-around) as described above for bars.
|
Plaids (Fig. 6
) consist of two superimposed gratings drifting in different directions; they normally yield an integrated percept of motion in the vector sum of their directions (type I plaid), or not (type II plaid), depending on the particular parameters used (Adelson and Movshon, 1982; Ferrera and Wilson, 1990). We used type I plaids to test for motion detection (moving versus non-moving) and direction discrimination (up versus down). The stimuli were implemented on the VSG system by spatially interleaving the images of the two component gratings (each a square wave of 1 c.p.d. with a contrast of 1.0 against a background of 10 cd/m2) oriented 135° apart and animated by manipulating the video look-up tables to achieve component speeds of 4 and 20°/s.
|
Motion in depth.
Random dot kinematograms depicting motion in depth (i.e. expanding versus contracting fields) were used to test for motion detection (expansion versus static), motion coherence detection (expansion versus 0% dot coherence) and direction discrimination (expansion versus contraction). The algorithm used to update the dot positions was used by Graziano and colleagues to study sensitivity to motion in depth of neurones in area MST of monkeys (Graziano et al., 1994
). Initially, each dot in the display was assigned to a random location, a direction corresponding to a radial trajectory, and a random age up to the maximum lifetime. In subsequent frames each dot was displaced in a radial direction by distance k x r, where k is a constant and r is the radial distance between the current location and the focus of expansion or contraction, and its age is incremented by the frame duration. Dots which reached their age limit were reassigned to a random location and their age was reset to zero. In the current displays values of k used were chosen to yield speeds of 4, 20, 32 or 64°/s at a radius of 10° from the focus of expansion or contraction. In our implementation, dot diameter was 0.25°, dot density 0.414 dots/deg2, and the dots were updated synchronously at a rate of 18.6 Hz. |
Second-order stimuli |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
Stimuli defined by static texture contrast were implemented on the VSG system (Figs 8 and 9
). These stimuli may be envisaged as a static textured surface viewed through a masking overlay with a single window (depicting a bar, 20° x 1°) or several small windows (depicting dots, 1° x 1° square) through which the textured surface was visible. The texture consisted of a black and white chequerboard (mean luminance 10 cd/m2, contrast 1.0) with an element size of 0.2° x 0.2°, and the overlay consisted of a uniform grey field whose luminance matched the space-averaged luminance of the textured surface. As the overlay was moved over the static texture it created the impression that the textured stimulus was moving. As long as the overlay was panned in steps equivalent to whole multiples of the size of the texture elements, there should have been be no first-order luminance cues associated with it.
|
|
Stimuli defined by dynamic texture contrast (Figs 8 and 9
) were implemented on the SVGA system. In these cases, the background consisted of an array of 0.2° x 0.2° elements, each allocated a luminance selected at random from a specified range of values, and the overlay consists of an identical array of elements, each allocated a luminance selected at random from a different range of values with the same mean (10 cd/m2). Thus, the bar or dots were differentiated from the background because of the different contrast ranges of their constituent elements. As before, animation was achieved by panning the overlay past the background in increments equal to whole multiples of the element size but, each time this occurred, the luminance of each element in the display was replaced with a new value selected at random from the appropriate range. This gave the impression of a flickering stimulus moving across a flickering background. The purpose of adding this noise was to mask any residual luminance-based cues for motion, such as might occur at relatively small spatial scales in the static texture display. This was implemented by manipulating the video look-up tables, with half the values in the 6-bit SVGA look-up table allocated to the background and half to the overlay. The luminance values were programmed to change every frame refresh (80 Hz, above the normal human flicker fusion rate) throughout the trial (including the interstimulus interval), the effect of which was to produce a broad band of temporal flicker frequencies across space and time (because only large random changes in the luminance of a texture element were visible). Second-order versions of the isolated bars and random dot kinematograms were used to test for motion detection (moving versus blank in the case of bars, and moving versus static in the case of kinematograms) and direction discrimination (up versus down).
Procedure
Tests were carried out in a diffusely lit room, with an ambient light level that just allowed reading. The subject sat at a table with his or her head supported by a chin rest and forehead restraint while gazing at a fixation spot. The fixation spot was a small white annulus fixed near the edge of a large sheet of black card with a window in it through which the stimuli were displayed on the VDU, at a viewing distance of 0.57 m and at an appropriate position in the visual field. This arrangement ensured that the visible portion of the VDU was confined entirely to the subject's field defect. The wall behind the VDU was covered with a black curtain which reflected very little light, but the intact hemifield was filled by a white background which reflected ambient light at an intensity of 5.0 cd/m2. G.Y. was tested monocularly (because that way it is easier to monitor fixation and control for light scatter than it is with binocular viewing), his left eye occluded by a light-tight, opaque patch; G.M. and S.P., both of whom were relatively inexperienced subjects, were tested binocularly because they found monocular tests tiring. Fixation was monitored with closed-circuit TV, and any trial in which subjects moved their eyes by any discernible amount was aborted. In practice this rarely occurred. Trials were initiated from, and subjects' verbal responses were relayed into, the keyboard of a microcomputer by an investigator who could see the image of the subject's eye(s) on the closed-circuit TV monitor but could not see the stimulus being displayed. A second experimenter sat behind and slightly to one side of the subject and observed every trial.
The ability to detect and discriminate movement in the various displays was measured with two-alternative forced choice tests. Trials consisted of two temporally contiguous intervals. Each consisted of a tone (200 ms), followed after 100 ms by the target or non-target system (which lasted 1000 ms except in the case of isolated bars; see above for details), followed after 1000 ms by the next interval, or a tone 1 octave higher indicating the end of the trial. The subject then indicated which interval (first or second) contained the target stimulus (detection), or in which interval the stimulus moved upwards as opposed to downwards or outwards as opposed to inwards (direction discrimination). Every condition was tested in the normal part of the subject's visual field before tests were carried out at the equivalent eccentricity in the field defect.
There were 40–100 trials for each stimulus condition (80 were planned, and in most cases achieved, for each condition). Stimuli were presented at four speeds (4, 20, 32 and 64°/s) in blocks of 40 trials, randomly interleaved unless otherwise indicated. Error bars in the graphs of results indicate 95% confidence limits, which take into account the influence of the actual number of trials on the estimate of performance.
|
Results |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
First-order motion
Translation
Bars: detection.
Subject G.Y. reliably detected the moving bar at a contrast of 0.75 when compared with a blank at each of the randomly interleaved speeds, though only just above the threshold for significance at the slowest (4°/s) and fastest (64°/s) speeds tested (Fig. 2
). Subjects G.M. and S.P. were tested only at maximum stimulus contrast. G.M. could detect the moving target reliably at all speeds except the slowest (4°/s), and he performed better when the trials were presented in blocks of the same speed (square symbols) as opposed to randomly interleaved (circles; although this was tested only at 32°/s). Subject S.P. was able to detect the moving target only at a speed of 32°/s (significant even after applying Bonferroni correction for multiple tests).
Bars: direction discrimination.
The results are shown in Fig. 2. G.Y.'s performance mirrored that for detection, i.e. he could discriminate direction of motion at speeds below 64°/s, though only just above the threshold for significance at the slowest speed (4°/s). G.M. was unable to determine the direction in which the bar moved at any speed when trials of different speeds were randomly interleaved, but when the trials were presented in blocks of the same speed he scored ~70% correct at maximum contrast and intermediate speeds. S.P. could barely discriminate motion direction at the fastest speed (not significant after applying Bonferroni correction), and could not do so at any other speed.
The speed–tuning curves for detection and discrimination in G.Y. and G.M. corresponded almost perfectly to those published previously for subject G.Y. by Barbur and colleagues (Barbur et al., 1993).
Random dot kinematograms: detection.
The results for all three subjects when discriminating between a moving, 100% coherent random dot kinematogram and a static one are shown in Fig. 3. At maximum contrast, G.Y.'s detection was perfect at every speed. However, at a contrast of 0.75, which was still about 2.0 log units above the contrast detection threshold measured in normal subjects, his detection fell to ~75% correct at speeds of 4 and 32°/s, and to just above 60% correct at 64°/s. Only at 20°/s did it reach 90% correct. At maximum contrast, subject G.M. performed just above chance at all speeds when trials of different speed were randomly interleaved, but performance was excellent when trials were presented in blocks of the same speed. Subject S.P. was tested only at maximum contrast with blocked presentations, and detected motion well at all speeds.
Subjects G.Y. and G.M. were also tested for their ability to detect motion in random dot kinematograms with dot lifetimes limited to 220 ms (Fig. 4). When discriminating moving stimuli from static ones in which the dots were replotted in random positions at the end of their lifetime, G.Y. could correctly identify the moving stimulus at all speeds except 4°/s, whereas G.M. could only identify the moving stimulus at 64°/s. When discriminating stimuli with 100% coherent dot movement from 0% coherent dot movement, neither subject could identify coherent movement correctly at any of the speeds tested.
|
Random dot kinematograms: direction discrimination.
With respect to the direction of motion of 100% coherent random dot kinematograms (Fig. 3
), none of the subjects performed significantly above chance at any of the speeds tested. In other words, even when they could detect the movement of a random dot kinematogram they were blind to its direction.
Gratings and plaids.
Figure 5 shows that both G.Y. and S.P. readily detected the movement of gratings, but failed to discriminate direction of movement at any of the speeds tested (four speeds for S.P., two for G.Y.).
G.Y. was also tested with plaids (Fig. 6) moving at speeds of 4 and 20°/s. As with gratings, he easily detected the moving stimulus, but could not discriminate its direction.
Motion in depth
Subjects G.Y. and G.M. were tested for their ability to detect and discriminate motion in depth using random dot kinematograms with limited dot lifetime. The results are shown in Fig. 7. When asked to discriminate between a kinematogram depicting expansion and a field of static dots, G.Y. could only do so reliably with expansion factors corresponding to maximum speeds of 32 and 64°/s at the edge of the stimulus. G.M. could not detect the movement of kinematograms with dots of 0.25° in diameter in the standard position in his field defect (circles), but he could do so reliably with 0.5° dots after the stimulus had been repositioned more centrally in the visual field (centred 20° to the left of the fixation point and 15° above it), and the maximum speed was 64°/s. Neither subject, however, could detect the difference between an expanding stimulus and one in which the dot motion was 0% coherent, nor could either of them discriminate between expansion and contraction.
|
Second-order motion
Subject G.Y. was also tested for his ability to detect and discriminate the direction of motion of second-order bars and random dot kinematograms defined by static or dynamic texture contrast.
Isolated bars
The results are plotted in Fig. 8. When the second-order moving bars were defined by static contrast (with a contrast range of 0.00–0.95 inside the bar against a uniform background), G.Y. was able to detect the moving stimulus at all speeds except the slowest (4°/s), and when asked to discriminate between upwards or downwards movement he could do so reliably (~80% correct) at intermediate speeds (16 and 32°/s). This pattern of results is very like that obtained with a first-order moving bar.
His performance was better when tested with bars defined by dynamic texture contrast. With a within-bar contrast range of 0–0.95 against a uniform background, he could both detect and discriminate the direction of the stimulus almost perfectly at any speed, despite the fact that he was unable to detect the motion at 4 and 64°/s, or to discriminate direction at 4°/s when the bar was defined by an identical amount of static texture contrast. With a within-bar contrast range of 0.0–0.95 set against a background contrast range of 0.0–0.20, he could detect the stimulus at almost any speed.
Random dot kinematograms
The results are shown in Fig. 9. With stimuli defined by static texture contrast (the moving dots being defined by a contrast range of 0.00–0.95 against uniform background), G.Y. could detect the moving stimulus at all speeds, though performance was better at 20°/s or above (
90% correct) than at 4°/s (65% correct), but he could not discriminate direction of motion at any speed. This is the same pattern of results as that obtained with first-order (luminance defined) kinematograms.
With stimuli defined by dynamic texture contrast, G.Y. could not detect the moving stimulus at any speed, even at the favourable contrast range of 0.00–0.95 against a uniform background. This is in striking contrast to his performance with first-order random dot kinematograms and second-order ones defined by static texture contrast.
The results are summarized in Table 1.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
Is blindsight motion blind?
Motion detection
All three subjects could detect every type of motion and discriminate its direction easily in the normal part of the visual field and at the eccentricity used in the field defect. Two of them, G.M. and G.Y., could also reliably detect isolated first-order moving bars, in their field defects, and were best at intermediate speeds (20 and 32°/s) (Fig. 1
). The results correspond almost perfectly to those obtained previously by Barbur and colleagues from subject G.Y. (Barbur et al., 1993).
Subject S.P. was relatively impaired at detecting isolated moving bars compared with the other two subjects. (She could only detect them reliably at 32°/s, but could not discriminate their direction at any speed.) The reason is not known, but it might be related to the extent of her lesion, which extended further anteriorly in the ventromedial occipital cortex, and may therefore have involved relatively more extrastriate visual cortex than in the other subjects. Evidence for the importance of the extrastriate cortex in blindsight comes from hemispherectomized patients. Although several early studies showed that such patients could discriminate visual targets in their scotoma, including moving ones, it now seems that their performance can be accounted for by a combination of artefacts, including response bias and light scatter (Campion et al., 1983; King et al., 1996; Barton and Sharpe, 1997; Faubert et al., 1999). The inability of hemispherectomized patients to respond explicitly to visual stimuli when such artefacts are carefully eliminated, even though implicit responses may persist (Tomaiuolo et al., 1997), suggests that the extrastriate cortex is necessary for mediating explicit visual responses (Azzopardi et al., 1996; King et al., 1996). In the case of S.P., one might have expected her ability to discriminate motion to be spared as her lesion appeared to have affected the ventral extrastriate cortex, which is specialized for analysing form (Ungerleider and Mishkin, 1982; Milner and Goodale, 1995) more than the dorsal extrastriate cortex (which includes area MT, specialized for motion processing), unless the dorsal stream contributes nothing at all to motion detection in the absence of the striate cortex. This is unlikely, given that neurones in area MT of monkeys with V1 lesions can respond selectively to the speed of moving targets, if not their direction (Azzopardi et al., 1998). Perhaps her lesion additionally affected the function of dorsal stream areas in a way that cannot be discerned from structural MRIs alone.
Attempts to assess motion detection by using a blank screen as a control for a moving bar confound the movement of the bar with its presence. However, the presence of the bar cannot, on its own, account entirely for G.Y.'s and G.M.'s performance, as their ability to detect the slowest bars (4°/s) was at, or close to, chance when the stimuli would have been present for the longest amount of time. Simply using a stationary bar instead of a blank screen would avoid the confound at the expense of introducing a new one, namely the markedly different temporal frequency spectra of the onset–offset of a static bar and a swept bar. This means that there is probably no adequate control that isolates the motion of a bar from its other properties.
All three subjects easily discriminated first-order, 100% coherent random dot kinematograms from static ones, with the exception of G.M. at 4°/s (Fig. 2). Flicker and movement are confounded in these stimuli, in the same way as they are when comparing moving and static bars. One attempt to ameliorate this was to compare 100% coherent random dot kinematograms with static displays with random relocation. Both G.Y. and G.M. could discriminate the moving stimulus from the static one (Fig. 4) (S.P. was not tested), but again it is impossible to equate the temporal frequency spectra of the two stimuli. Perhaps the best way around the problem is to compare 100% coherent random dot kinematograms with 0% coherent ones, which should have very similar, if not identical, temporal frequency spectra, but as both stimuli contain the same number of elements moving at the same speed, this should reveal more about the ability to discriminate speed (a directionless quantity) from velocity (a vector) than about the ability to discriminate moving from non-moving stimuli. In practice, neither G.Y. or G.M. could discriminate between them (Fig. 4).
Sensitivity to flicker seems to be the most important factor underlying motion detection in the scotoma. This is borne out by two experimental observations. The first is that G.Y.'s sensitivity to moving bars peaked at a speed of 20°/s (Fig. 1), corresponding to a fundamental temporal frequency of 10 Hz for a 1° wide bar swept past a point on the screen at that speed, which in turn corresponds to the peak of the relatively restricted range of temporal frequency sensitivity in his scotoma (Barbur et al., 1994). Secondly, G.Y. was able at all speeds to detect second-order, moving bars when they were defined by dynamic contrast with broad-band temporal frequency spectra, but not when they were defined by static contrast (Fig. 8), even though the average difference in contrast between the target and background was much lower in the former stimulus.
Despite the difficulty of separating the contributions of movement and flicker to the detection of moving targets, it is important to know that both moving bars and random dot kinematograms are easily detectable in the scotoma, whichever cue might be used.
Direction discrimination
G.M. and G.Y. could reliably discriminate the direction of isolated first-order moving bars, and were best at intermediate speeds (20 and 32°/s), which reflects the detectability of the stimuli (Fig. 2). Again, the results correspond almost perfectly to those obtained previously by Barbur and colleagues with G.Y. (Barbur et al., 1993). But in stark contrast, neither subject could discriminate the direction of motion of 100% coherent random dot kinematograms (the easiest possible condition), despite having no difficulty in detecting the presence of movement in the stimuli (Fig. 3). The same was true for G.Y. when tested with gratings and plaids (Figs 5 and 6), and for both subjects when tested with random dot kinematograms depicting motion in depth (Fig. 7). These results are in accordance with previously published findings (King et al., 1996; Barton and Sharpe, 1997).
The most obvious difference between isolated moving bars and the remaining stimuli is that there is a distinct shift in the centre of gravity of the stimulus associated with a bar, but not with the other stimuli, which leads to the suspicion that scattered light might be the basis for direction discrimination with bars. This can be discounted because G.Y. could just as easily discriminate the direction of motion of second-order bars which had no luminance cues associated with them (Fig. 8). Unfortunately, neither G.M. nor S.P. could be tested with second-order motion stimuli in the time available, so that light scatter cannot be excluded as the basis of their performance.
The fact that neither G.M. nor G.Y. could discriminate the direction of motion depicted by random dot kinematograms or gratings or plaids suggests, contrary to common belief, that motion processing is severely impaired in the scotoma. This is consistent with the fact that neurones in cortical area MT of monkeys with striate cortex lesions, which are sensitive to the direction of moving bars presented in the scotoma (Rodman et al., 1989; Girard et al., 1992), are nevertheless insensitive to the direction of motion of random dot kinematograms, gratings and plaids (Azzopardi et al., 1998; Fallah et al., 1998). It explains much of the discrepancy in the literature on motion processing in blindsight, as the clearest evidence in favour of motion processing was obtained with isolated bars and spots as stimuli (Barbur et al., 1993; Weiskrantz et al., 1995), and the clearest evidence against motion processing used random dot kinematograms (Barton and Sharpe, 1997). The exceptions, in which direction discrimination was reported with random dot kinematograms and gratings, could be due to artefacts [some of which are discussed elsewhere (Cowey and Azzopardi, 2000)].
The nature of the impairment
Motion perception is thought of as being mediated by a combination of low-level spatiotemporal (Fourier) mechanisms and higher-level (non-Fourier) feature-tracking mechanisms (for review, see Clifford and Vaina, 1999). Possible explanations of the dissociation between the ability to discriminate the direction of motion of isolated bars and the inability to discriminate the direction of motion of more complex stimuli can be classified accordingly.
One kind of explanation is that motion discrimination in the scotoma is mediated by somewhat impaired spatiotemporal (Fourier) mechanisms. For example, the spatiotemporal analysers might adapt abnormally quickly. Individual analysers would not adapt much as a result of a single sweep of a bar, but would adapt rapidly to the point of quiescence when subjected to a stream of dots (or bars, in the case of a grating) swept through their receptive fields in rapid succession. However, this would not be consistent with unpublished observations of the responses of MT neurones in monkeys with striate cortex lesions, which do not seem to adapt or habituate significantly in the course of a 1s presentation of a random dot kinematogram (P. Azzopardi, M. Fallah, C. G. Gross and H. R. Rodman, unpublished results).
Another possibility is that low-level spatiotemporal mechanisms are completely disabled as result of damaging the striate cortex, and motion perception is therefore completely reliant on feature-tracking mechanisms. Failure to discriminate the direction of motion of complex stimuli could then be accounted for by a failure to solve the correspondence problem (i.e. correctly assigning each feature in a frame to the corresponding feature in the previous frame), whereas the ability to discriminate the direction of a moving bar could be inferred from a change in its centre of gravity by feature-tracking mechanisms. This explanation is more consistent with the data presented in this paper, and experiments are currently under way to test it further.
Another aspect of the impairment concerns the question of whether or not patients are aware of moving stimuli in the field defect, i.e. whether or not they have blindsight for motion. Although this was not the question we set out to address, it deserves some comment as the issue is controversial and apparently unresolved. The impression created by the literature is that cortically blind patients are more likely to report being aware of moving stimuli than static ones (Riddoch, 1917; Holmes, 1918; Weiskrantz, 1990; Barbur et al., 1993; Weiskrantz et al., 1995; Zeki and ffytche, 1998). One explanation could be that, without first equating moving and non-moving targets for detectability, differences in the frequency of reported awareness could merely reflect differences in salience between the stimuli (Holmes, 1918). Recently, however, we used signal detection methods to compare G.Y.'s detection of static targets with his detection of moving targets from the same range of detectability (d'), and found a qualitative difference between his unbiased sensitivity to static targets and to moving targets, which meant that response bias was more likely to account for blindsight-like dissociations between yes–no and forced-choice detection of moving stimuli than of static patterns (Azzopardi and Cowey, 1997, 1998). The confusion in the literature over whether or not G.Y. is aware of moving stimuli could therefore have arisen simply because previous studies used potentially biased measures of awareness (such as per cent correct). The main difference between static and moving targets may lie in the extent to which their representations allow patients with striate cortex lesions to set consistent decision criteria in yes–no tasks, which include answering the question `Were you aware of . . .?' (Azzopardi and Cowey, 2000b).
Implications for the functional architecture of motion perception
It has been suggested that motion perception is mediated by two dissociable pathways, namely a route from the retina to cortical area MT via subcortical nuclei that bypasses striate cortex and is specialized for processing fast motion (>6°/s), and an indirect cortical route via the striate cortex (V1) for processing slow motion (<6°/s) (ffytche et al., 1995; Zeki, 1998). This hypothesis, which is called dynamic parallelism, is founded on a double dissociation in which patients with a lesion affecting the striate cortex, such as G.Y., are able to detect and discriminate reliably the direction of isolated bars moving inside the scotoma at speeds greater than ~6°/s but are unable to do so at lower speeds (Barbur et al., 1993), whereas a patient with bilateral lesions of MT (patient L.M.) can detect and discriminate the direction of movement at speeds below 6°/s but not at higher speeds (Zihl et al., 1983; Hess et al., 1989).
There is no doubt that cortical area MT is important for the perception of motion. In monkeys, >90% of neurones in MT are sensitive to direction of motion (Zeki, 1974; Maunsell and van Essen, 1983; Albright, 1984), the destruction of area MT can cause impairments in direction discrimination tasks (Newsome and Pare, 1988; Pasternak and Merigan, 1994), the response rates of neurones in MT covary with the monkey's choice of direction during direction discrimination tasks (Britten et al., 1992), and stimulation of MT neurones can influence the monkey's perception of direction during direction discrimination tasks (Salzman et al., 1992). In humans, bilateral destruction of a cortical area homologous with MT (Clarke and Miklossy, 1990) can cause selective impairments of motion perception (Zihl et al., 1983; Hess et al., 1989); activation of area MT's homologue assessed with PET and functional MRI is correlated with the presentation of moving stimuli viewed passively or in the context of motion discrimination tasks (Zeki et al., 1991; Watson et al., 1993; Tootell et al., 1995); and transient inactivation of MT's homologue by transcranial magnetic stimulation impairs motion perception in normal subjects (Beckers and Homberg, 1992) and can produce illusory moving phosphenes (Stewart et al., 1999). However, there is doubt as to whether there exists a sufficient functional pathway to MT that bypasses the striate cortex. Rodman and colleagues recorded from neurones in area MT of monkeys with striate cortex lesions and found that ~50% of neurones retained directionally selective responses to isolated moving bars swept through their receptive fields (though their sensitivity was reduced) (Rodman et al., 1989). These responses were completely abolished by a second lesion of the superior colliculus (Rodman et al., 1990), indicating that MT must have been activated through a pathway via the superior colliculus and pulvinar. This does not necessarily mean that this pathway is of any functional relevance in the intact monkey brain (Gross, 1991). Further, there is now some doubt as to whether the responses recorded by Rodman and colleagues (Rodman et al., 1989) reflect unimpaired motion processing, as neurones in MT with receptive field in the scotoma do not respond selectively to the direction of motion depicted by 100% coherent random dot kinematograms, at either slow (4°/s) or fast (20°/s) speeds (Azzopardi et al., 1998).
Another line of evidence for dynamic parallelism comes from studying the latency of responses to moving stimuli in area MT. ffytche and colleagues, using a combination of EEG and MEG (magnetoencephalography) in normal human subjects, found that responses to slow stimuli (<6°/s) were evoked in striate cortex before V5, whereas responses to fast stimuli (22°/s) were evoked in V5 before the striate cortex, implying that visual information is routed to MT via one of two parallel routes (one through the striate cortex and one bypassing it). But this finding is controversial, as it could not be replicated with the more sensitive method of MEG (Anderson et al., 1996). Also, if ffytche and colleagues (ffytche et al., 1995) are correct and responses in area MT after striate cortex lesions reflect the normal function of the retinocolliculopulvinar–MT pathway, then short-latency responses in MT to fast stimuli should be preserved and long-latency responses to slow stimuli should be abolished by striate cortex lesions, as found in G.Y. by ffytche and colleagues (ffytche et al., 1996). This is also controversial, as at least three independent groups have been unable to replicate it (Holliday et al., 1997; Benson et al., 1999; Rao et al., 1999), and it is not consistent with the fact that latencies of neurones in area MT to slow (4°/s) and fast (20°/s) moving stimuli are not significantly different in monkeys after striate cortex lesions (Azzopardi et al., 1999).
So far the only undisputed evidence for dynamic parallelism is the double dissociation between patients like G.Y., with lesions affecting the striate cortex, and those like L.M., with lesions affecting MT/V5. There is no doubt that both G.Y. and G.M. are more sensitive to and better able to discriminate the direction of isolated bars moving through the scotoma at 20°/s as opposed to 4°/s (Barbur et al., 1993) (see also Fig. 2), but their inability to discriminate the direction of random dot kinematograms (or gratings) at either speed gives cause for concern as it suggests that motion processing is severely impaired after striate cortex lesions. If this is the case, then there are no grounds for inferring that there is an alternative route to the extrastriate visual cortex that supports motion perception adequately.
|
Acknowledgments |
|---|
This work was supported by Medical Research Council grant G971/387/B.
|
References |
|---|
|
TopAbstractIntroductionMethodsSubjectsSecond-order stimuliResultsDiscussionReferences |
|---|
Adelson EH, Movshon JA. Phenomenal coherence of moving visual patterns. Nature 1982; 300: 523–5.[Medline]
Albright TD. Direction and orientation selectivity of neurons in visual area MT of the macaque. J Neurophysiol 1984; 52: 1106–30.
Anderson SJ, Holliday IE, Singh KD, Harding GF. Localization and functional analysis of human cortical area V5 using magneto-encephalography. Proc R Soc Lond B Biol Sci 1996; 263: 423–31.[Medline]
Azzopardi P, Cowey A. Is blindsight like normal, near-threshold vision? Proc Natl Acad Sci USA 1997; 94: 14190–4.
Azzopardi P, Cowey A. Blindsight and visual awareness. Conscious Cogn 1998; 7: 292–311.[Web of Science][Medline]
Azzopardi P, Cowey A. Why blindsight is blind. In: de Gelder B, de Haan E, Heywood C, editors. Varieties of unconscious processing: new findings and new models. Oxford: Oxford University Press. In press 2000.
Azzopardi P, King SM, Cowey A. Human pattern electroretinograms after complete cerebral hemispherectomy [abstract]. Soc Neurosci Abstr 1996; 22: 399.
Azzopardi P, Fallah M, Gross, CG, Rodman HR. Responses of neurons in visual areas MT and MST after lesions of striate cortex in macaque monkeys [abstract]. Soc Neurosci Abstr 1998; 24: 648.
Azzopardi P, Fallah M, Gross CG. Response latencies of neurons in areas MT and MST in monkeys with striate cortex lesions. Invest Ophthalmol Vis Sci 1999; 40: S640.
Barbur JL, Ruddock KH, Waterfield VA. Human visual responses in the absence of the geniculo-calcarine projection. Brain 1980; 103: 905–28.
Barbur JL, Watson JD, Frackowiak RS, Zeki S. Conscious visual perception without V1. Brain 1993; 116: 1293–302.
Barbur JL, Harlow AJ, Weiskrantz L. Spatial and temporal response properties of residual vision in a case of hemianopia. Philos Trans R Soc Lond B Biol Sci 1994; 343: 157–66.
Barton JJ, Sharpe JA. Motion direction discrimination in blind hemifields. Ann Neurol 1997; 41: 255–64.[Web of Science][Medline]
Beckers G, Homberg V. Cerebral visual motion blindness: transitory akinetopsia induced by transcranial magnetic stimulation of human area V5. Proc R Soc Lond B Biol Sci 1992; 249: 173–8.[Medline]
Benson PJ, Guo K, Blakemore C. Direction discrimination of moving gratings and plaids and coherence in dot displays without primary visual cortex (V1). Eur J Neurosci 1998; 10: 3767–72.[Web of Science][Medline]
Benson PJ, Guo K, Hardiman MJ. Cortical evoked potentials due to motion contrast in the blind hemifield. Neuroreport 1999; 10: 3595–600.[Web of Science][Medline]
Blythe IM, Bromley JM, Kennard C, Ruddock KH. Visual discrimination of target displacement remains after damage to the striate cortex in humans. Nature 1986; 320: 619–21.[Medline]
Blythe IM, Kennard C, Ruddock KH. Residual vision in patients with retrogeniculate lesions of the visual pathways. Brain 1987; 110: 887–905.
Britten KH, Shadlen MN, Newsome WT, Movshon JA. The analysis of visual motion: a comparison of neuronal and psychophysical performance. J Neurosci 1992; 12: 4745–65.[Abstract]
Bullier J, Girard P, Salin P-A. The role of area 17 in the transfer of information to extrastriate visual cortex. In: Peters A, Rockland KS, editors. Cerebral cortex, Vol. 10. New York: Plenum Press; 1994. p. 301–30.
Campion J, Latto R, Smith YM. Is blindsight an effect of scattered light, spared cortex and near-threshold vision? Behav Brain Sci 1983; 6: 423–86.[Web of Science]
Cavanagh P, Mather G. Motion: the long and the short of it. Spat Vis 1989; 4: 103–29.[Medline]
Chubb C, Sperling G. Drift-balanced random stimuli: a general basis for studying non-Fourier motion perception. J Opt Soc Amer A 1988; 5: 1986–2007.
Clarke S, Miklossy J. Occipital cortex in man: organization of callosal connections, related myelo- and cytoarchitecture, and putative boundaries of functional visual areas. J Comp Neurol 1990; 298: 188–214.[Web of Science][Medline]
Clifford CW, Vaina LM. A computational model of selective deficits in first and second-order motion processing. Vision Res 1999; 39: 113–30.[Web of Science][Medline]
Cowey A, Azzopardi P. Is blindsight motion-blind? In: de Gelder B, de Haan E, Heywood C, editors. Varieties of unconscious processing: new findings and new models. Oxford: Oxford University Press. In press 2000.
Cowey A, Stoerig P. The neurobiology of blindsight. [Review]. Trends Neurosci 1991; 14: 140–5.[Web of Science][Medline]
Dubner R, Zeki SM. Response properties and receptive fields of cells in an anatomically defined region of the superior temporal sulcus in the monkey. Brain Res 1971; 35: 528–32.[Web of Science][Medline]
Fallah M, Azzopardi P, Gross CG, Rodman HR. Visual motion sensitivity of extrastriate neurons in macaques with long standing lesions of striate cortex. J Cogn Neurosci 1998; 10: S14.
Faubert J, Diaconu V, Ptito M, Ptito A. Residual vision in the blind field of hemidecorticated humans predicted by a diffusion scatter model and selective spectral absorption of the human eye. Vision Res 1999; 39: 149–57.[Web of Science][Medline]
Ferrera VP, Wilson HR. Perceived direction of moving two-dimensional patterns. Vision Res 1990; 30: 273–87.[Web of Science][Medline]
ffytche DH, Guy CN, Zeki S. The parallel visual motion inputs into areas V1 and V5 of human cerebral cortex. Brain 1995; 118: 1375–94.
ffytche DH, Guy CN, Zeki S. Motion specific responses from a blind hemifield. Brain 1996; 119: 1971–82.
Girard P, Salin PA, Bullier J. Visual activity in areas V3a and V3 during reversible inactivation of area V1 in the macaque monkey. J Neurophysiol 1991; 66: 1493–503.
Girard P, Salin PA, Bullier J. Response selectivity of neurons in area MT of the macaque monkey during reversible inactivation of area V1. J Neurophysiol 1992; 67: 1437–46.
Goldberg ME, Wurtz RH. Activity of superior colliculus in behaving monkey. I. Visual receptive fields of single neurons. J Neurophysiol 1972; 35: 542–59.
Graziano MS, Andersen RA, Snowden RJ. Tuning of MST neurons to spiral motions. J Neurosci 1994; 14: 54–67.[Abstract]
Gross CG. Contribution of striate cortex and the superior colliculus to visual function in area MT, the superior temporal polysensory area and the inferior temporal cortex. Neuropsychologia 1991; 29: 497–515.[Web of Science][Medline]
Hess RH, Baker CL Jr, Zihl J. The `motion-blind' patient: low-level spatial and temporal filters. J Neurosci 1989; 9: 1628–40.[Abstract]
Holliday IE, Anderson SJ, Harding GF. Magnetoencephalographic evidence for non-geniculostriate visual input to human cortical area V5. Neuropsychologia 1997; 35: 1139–46.[Web of Science][Medline]
Holmes G. Disturbances of vision by cerebral lesions. Br J Ophthalmol 1918; 2: 353–84.
Holmes G. The organization of visual cortex in man. Proc R Soc Lond 1945; B132: 348–61.
Kentridge RW, Heywood CA, Weiskrantz L. Residual vision in multiple retinal locations within a scotoma: implications for blindsight. J Cogn Neurosci 1997; 9: 191–202.[Web of Science]
King SM, Azzopardi P, Cowey A, Oxbury J, Oxbury S. The role of light scatter in the residual visual sensitivity of patients with complete cerebral hemispherectomy. Vis Neurosci 1996; 13: 1–13.[Web of Science][Medline]
Maunsell JH, Van Essen DC. Functional properties of neurons in middle temporal visual area of the macaque monkey: I. Selectivity for stimulus direction, speed and orientation. J Neurophysiol 1983; 49: 1127–47.
Mestre DR, Brouchon M, Ceccaldi M, Poncet M. Perception of optical flow in cortical blindness: a case report. Neuropsychologia 1992; 30: 783–95.[Web of Science][Medline]
Milner DA, Goodale MA. The visual brain in action. Oxford: Oxford University Press; 1995.
Mohler CW, Wurtz RH. Role of striate cortex and superior colliculus in visual guidance of saccadic eye movements in monkeys. J Neurophysiol 1977; 40: 74–94.
Morland AB, Jones SR, Finlay AL, Deyzac E, Le S, Kemp S. Visual perception of motion, luminance and colour in a human hemianope. Brain 1999; 122: 1183–98.
Newsome WT, Pare EB. A selective impairment of motion perception following lesions of the middle temporal visual area (MT). J Neurosci 1988; 8: 2201–11.[Abstract]
Pasternak T, Merigan WH. Motion perception following lesions of the superior temporal sulcus in the monkey. Cerebral Cortex 1994; 4: 247–59.
Perenin MT. Discrimination of motion direction in perimetrically blind fields. Neuroreport 1991; 2: 397–400.[Web of Science][Medline]
Pöppel E, Held R, Frost D. Residual visual function after brain wounds involving the central visual pathways in man [letter]. Nature 1973; 243: 295–6.[Medline]
Rao A, Nobre AC, Cowey A. Comparison of visual evoked potentials in a patient with a V1 lesion and in normal subjects [abstract]. Soc Neurosci Abstr 1999; 25: 1931.
Riddoch G. Dissociation of visual perceptions due to occipital injuries, with especial reference to appreciation of movement. Brain 1917; 40: 15–57.
Rodman HR, Gross CG, Albright TD. Afferent basis of visual response properties in area MT of the macaque. I. Effects of striate cortex removal. J Neurosci 1989; 9: 2033–50.[Abstract]
Rodman HR, Gross CG, Albright TD. Afferent basis of visual response properties in area MT of the macaque. II. Effects of superior colliculus removal. J Neurosci 1990; 10: 1154–64.[Abstract]
Salzman CD, Murasugi CM, Britten KH, Newsome WT. Microstimulation in visual area MT: effects on direction discrimination performance. J Neurosci 1992; 12: 2331–55.[Abstract]
Stewart L, Battelli L, Walsh V, Cowey A. Motion perception and perceptual learning studied by magnetic stimulation. Electroencephalogr Clin Neurophysiol Suppl 1999; 51: 334–50.[Medline]
Stoerig P, Cowey A. Blindsight in man and monkey. [Review]. Brain 1997; 120: 535–59.
Tomaiuolo F, Ptito M, Marzi CA, Paus T, Ptito A. Blindsight in hemispherectomized patients as revealed by spatial summation across the vertical meridian. Brain 1997; 120: 795–803.
Tootell RB, Reppas JB, Kwong KK, Malach R, Born RT, Brady TJ, et al. Functional analysis of human MT and related visual cortical areas using magnetic resonance imaging. J Neurosci 1995; 15: 3215–30.[Abstract]
Ungerleider LG, Mishkin M. Two cortical visual systems. In: Ingle DJ, Goodale MA, Mansfield RJW, editors. Analysis of visual behaviour. Cambridge (MA): MIT Press; 1982. p. 549–86.
Watson JD, Myers R, Frackowiak RS, Hajnal JV, Woods RP, Mazziotta JC, et al. Area V5 of the human brain: evidence from a combined study using positron emission tomography and magnetic resonance imaging. Cerebral Cortex 1993; 3: 79–94.
Weiskrantz L. Contour discrimination in a young monkey with striate cortex ablation. Neuropsychologia 1963; 1: 145–64.[Web of Science]
Weiskrantz L. Blindsight: a case study and its implications. Rev. ed. Oxford: Clarendon Press; 1990.
Weiskrantz L, Warrington EK, Sanders MD, Marshall J. Visual capacity in the hemianopic field following a restricted occipital ablation. Brain 1974; 97: 709–28.
Weiskrantz L, Harlow A, Barbur JL. Factors affecting visual sensitivity in a hemianopic subject. Brain 1991; 114: 2269–82.
Weiskrantz L, Barbur JL, Sahraie A. Parameters affecting conscious versus unconscious visual discrimination with damage to the visual cortex (V1). Proc Natl Acad Sci USA 1995; 92: 6122–6.
Zeki SM. Functional organization of a visual area in the posterior bank of the superior temporal sulcus of the rhesus monkey. J Physiol (Lond) 1974; 236: 549–73.
Zeki S. Parallel processing, asynchronous perception, and a distributed system of consciousness in vision. Neuroscientist 1998; 4: 365–72.
Zeki S, ffytche DH. The Riddoch syndrome: insights into the neurobiology of conscious vision. [Review]. Brain 1998; 121: 25–45.
Zeki S, Watson JD, Lueck CJ, Friston KJ, Kennard C, Frackowiak RS. A direct demonstration of functional specialization in human visual cortex. J Neurosci 1991; 11: 641–9.[Abstract]
Zihl J, von Cramon D, Mai N. Selective disturbance of movement vision after bilateral brain damage. Brain 1983; 106: 313–40.
Received March 13, 2000. Revised June 26, 2000. Second revision on September 11, 2000. Accepted September 11, 2000.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. R. Huxlin, T. Martin, K. Kelly, M. Riley, D. I. Friedman, W. S. Burgin, and M. Hayhoe Perceptual Relearning of Complex Visual Motion after V1 Damage in Humans J. Neurosci., April 1, 2009; 29(13): 3981 - 3991. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Bridge, O. Thomas, S. Jbabdi, and A. Cowey Changes in connectivity after visual cortical brain damage underlie altered visual function Brain, June 1, 2008; 131(6): 1433 - 1444. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Azzopardi, S. M. King, and A. Cowey Pattern electroretinograms after cerebral hemispherectomy Brain, June 1, 2001; 124(6): 1228 - 1240. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||