Skip to main content
Journal ArticleOPEN ACCESS

Modelling visual neglect: Computational insights into conscious perception

PLoS ONE (2010) 5(6)
DOI: 10.1371/journal.pone.0011128
10Citations
Citations of this article
59Readers
Mendeley users who have this article in their library.

Abstract

Background: Visual neglect is an attentional deficit typically resulting from parietal cortex lesion and sometimes frontal lesion. Patients fail to attend to objects and events in the visual hemifield contralateral to their lesion during visual search. Methodology/Principal Finding: The aim of this work was to examine the effects of parietal and frontal lesion in an existing computational model of visual attention and search and simulate visual search behaviour under lesion conditions. We find that unilateral parietal lesion in this model leads to symptoms of visual neglect in simulated search scan paths, including an inhibition of return (IOR) deficit, while frontal lesion leads to milder neglect and to more severe deficits in IOR and perseveration in the scan path. During simulations of search under unilateral parietal lesion, the model's extrastriate ventral stream area exhibits lower activity for stimuli in the neglected hemifield compared to that for stimuli in the normally perceived hemifield. This could represent a computational correlate of differences observed in neuroimaging for unconscious versus conscious perception following parietal lesion. Conclusions/Significance: Our results lead to the prediction, supported by effective connectivity evidence, that connections between the dorsal and ventral visual streams may be an important factor in the explanation of perceptual deficits in parietal lesion patients and of conscious perception in general. © 2010 Lanyon, Denham.

Figures

  • Figure 1. Examples from neglect patients. a. Example scan path from patient G.K. who has infarction of the right inferior parietal lobe but has sparing of the frontal lobe [1]. The search task was to find letter Ts amongst distractor Ls. There is profound neglect of the left side of the array and many re-fixations on the right. Figure from Husain et al. [1] ‘Impaired spatial working memory across saccades contributes to abnormal search in parietal neglect’ Brain (2001), 124, 941–952, by permission of Oxford University Press. b. Visual targets successfully cancelled during a cancellation task involving random shapes by Weintraub and Mesulam [52]. This patient has a large right parieto-temporal lesion and dramatically neglects the left hemifield. Reproduced from ‘Visual hemispatial inattention: stimulus parameters and exploratory strategies’ Weintraub and Mesulam [52] Journal of Neurology, Neurosurgery and Psychiatry, 51(12), 1481–8, (1988) with permission from BMJ Publishing Group Ltd. c. Visual targets successfully cancelled during a cancellation task involving randomly positioned letters by Weintraub and Mesulam [52]. This patient has a right superior frontal infarct and neglects the left hemifield. Reproduced from ‘Visual hemispatial inattention: stimulus parameters and exploratory strategies’ Weintraub and Mesulam [52] Journal of Neurology, Neurosurgery and Psychiatry, 51(12), 1481–8, (1988) with permission from BMJ Publishing Group Ltd. doi:10.1371/journal.pone.0011128.g001
  • Figure 2. Model Architecture. Schematic of the modules within the system. Note that the actual number of cells in V1, V4 and the parietal module are more than shown, and inhibitory interneurons are not shown. For further details see Lanyon and Denham [8,11]. doi:10.1371/journal.pone.0011128.g002
  • Figure 3. Simulated scan path following parietal lesion without leftward reset. A typical scan path obtained when the parietal module is unilaterally lesioned under overt attention and no leftward reset back into the scene is present at the rightmost border. The first fixation is placed at the centre of the image. From there the scan path is attracted to the right and becomes unable to re-orient away from the rightmost border of the scene. doi:10.1371/journal.pone.0011128.g003
  • Figure 4. Simulated scan paths following parietal lesion. a. A scan path obtained using the intact version of the model under overt attention. The target object is a red bar; hence most fixations land near red objects rather than near green objects or in blank regions of the scene. b. Scan path produced from overt attention when the parietal module is unilaterally lesioned. Severe symptoms of hemineglect are present in the scan path. This shows that visual hemineglect is produced when the parietal region that encodes stimuli in a retinotopic frame of reference is unilaterally lesioned and overt eye movements are made. This figure may be compared to the parietal patient behaviours shown in figure 1a and b. c. Attentional scanning movements produced by the intact model using covert attention. d. Symptoms of visual neglect in covert scanning produced when the model’s parietal module was unilaterally lesioned. This shows that visual hemineglect is produced when the parietal region that encodes stimuli in a retinotopic frame of reference is unilaterally lesioned and covert attention is deployed. This figure may be compared to the parietal patient behaviours shown in figure 1a and b. doi:10.1371/journal.pone.0011128.g004
  • Figure 5. Attentional capture by the left hemifield when there is a gradient of impairment. a. An example of covert scanning under a gradient lesion when the parietal module was lesioned to reflect a gradient of impairment. The impairment is strongest at the left and improves towards the centre of the left hemifield. At the leftmost periphery the associated parietal neurons are 7% impaired. Across the left hemisphere impairment decreases in a linear fashion so that neurons at the centre and those in the right hemifield are unimpaired. Whilst most red (target colour) stimuli are attended in the right hemifield and the more central region of the left hemifield, the impairment is greatest towards the far left and stimuli here are neglected. b. The same simulation as (a) but using overt attention. c. Shows that attention is more likely to be captured by stimuli in the left hemifield when the gradient of impairment is less steep, i.e. the effect of the lesion is less extensive. Mean percentages from 10 separate simulations, each containing 100 fixations and using covert attention, are shown. Effects saturate (i.e. the hemifield is completely neglected) at 20% lesion at the far left of the hemifield. doi:10.1371/journal.pone.0011128.g005
  • Figure 6. Effect of Parietal Lesion On V4 Activity. When presented with a scene containing a single red vertical bar in each hemifield, the responses of V4 cells that are selective for red or vertical features are shown. Responses from green and vertical selective cells are not shown since these were near baseline, due to lack of relevant stimuli in the scene. The plots on the left relate to cells that have receptive fields positioned left of the vertical meridian, in the neglected hemifield. The plots on the right are for cells with receptive fields in the normal right hemifield. Cells selective for the same feature have the same response properties except the position of their receptive fields. Activity for the stimulus in the neglected hemifield is reduced compared to that for the identical stimulus in the normal hemifield. doi:10.1371/journal.pone.0011128.g006
  • Figure 7. Frontal Lesion and Re-visiting Locations in the Scan Path. a. When the frontal module is unilaterally lesioned, symptoms of neglect are present in the scan path (overt attention shown here). This shows that a lesion in an area representing stimuli in a scene(or body)–based frame of reference can produce symptoms of neglect. This figure may be compared to the frontal neglect patient behaviour shown in figure 1c. b. A comparison of the percentage of fixations in the left hemifield following unilateral parietal versus unilateral frontal lesion in this model. The plot shows the mean percentages from 10 separate simulations, each containing 50 shifts of covert attention (values were similar under overt attention). Although the left hemifield is typically neglected following frontal lesion, some fixations are placed in this hemifield. Hence, compared to that following parietal lesion, neglect is less severe with frontal lesion in this model. c. When the frontal module is completely lesioned the scan path has difficulty exploring the scene and perseverates in one area. An overt attention simulation is shown here but similar effects are produced under conditions of covert attention. This is due to the lack of novelty bias in the system. d. Shows the effect of lesion on re-visiting of locations under covert attention. Locations are more frequently re-visited under conditions of parietal lesion than in normal conditions. However, lesion of the frontal bias in our model causes the greatest increase in re-visiting/re-fixation. The plot shows the mean percentages of re-visited locations from 10 separate simulations, each containing 50 shifts of attention. doi:10.1371/journal.pone.0011128.g007
  • Figure 8. Re-visiting Locations in the Scan Path Over Time. Shows the numbers of re-fixations (or simply re-visiting the same location during covert attentional scanning) that occur at each time lag since that location was first visited under (a) normal conditions, (b) unilateral parietal lesion, (c) unilateral frontal lesion, (d) bilateral frontal lesion. Similar to that reported by Mannan et al. [1,5], immediate re-fixations i.e. those occurring at the subsequent fixation (time lag of 1) have been removed. Whereas parietal lesion causes an increase in re-fixation only at longer time lags, frontal lesion causes re-fixation increases and short and longer time lags. doi:10.1371/journal.pone.0011128.g008

References Powered by Scopus

This article is free to access.

Effects of parietal injury on covert orienting of attention

Journal of Neuroscience
1842Citations
258Readers

Space and attention in parietal cortex

Annual Review of Neuroscience
1333Citations
1081Readers
Get full text
View more at Scopus

Cited by Powered by Scopus

View more at Scopus

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Cite

CITATION STYLE

APA

Lanyon, L. J., & Denham, S. L. (2010). Modelling visual neglect: Computational insights into conscious perception. PLoS ONE, 5(6). https://doi.org/10.1371/journal.pone.0011128

Readers over time

‘10‘11‘12‘13‘14‘15‘16‘17‘18‘19‘20‘21‘22‘24036912

Readers' Seniority

Tooltip

PhD / Post grad / Masters / Doc 22

49%

Researcher 11

24%

Professor / Associate Prof. 8

18%

Lecturer / Post doc 4

9%

Readers' Discipline

Tooltip

Psychology 24

62%

Neuroscience 8

21%

Medicine and Dentistry 4

10%

Computer Science 3

8%

Save time finding and organizing research with Mendeley

Sign up for free
0