Extinction studies

xegaxezu's version from 2015-06-07 15:29


Neuroanatomy of Extinction
Chechlacz et al (2013):
Study used voxel-based morphology (VBM) to analyse lesions in patients with visual extinction and compared them to patients without extinction. They found that TPJ damage was associated with extinction, which showed a possible overlap with neglect syndrome. However, several other areas were damaged in visual extinction patients which are not damaged in neglect, including the angular and middle occipital gyri, the insula, the superior temporal sulcus and the IPL. They also analysed lesions to white matter tracts in patients by using cytoarchitectonic probabilistic maps to work out the extent of the damage to the superior longitudinal fasciculus (SLF) in patients with visual extinction. They analysed damage to the SLF as it is thought to be part of the network for directing spatial attention. There are several different components of the SLF which connect different cortical areas: SLFI links the superior and medial parietal cortex with premotor areas, while SLFII projects from the IPL and occipito-parietal areas to the dlPFC and the SLFIII projects from rostral IPL to ventral premotor areas. The researchers found that 32% of the voxels which were attributed to the right SLF were in the same area as the white matter lesions associated with visual extinction. As damage to the SLF is present in visual extinction, and the SLF is important for connecting posterior parietal areas such as the TPJ and IPL to other areas of the brain, it suggests that these areas need to be damaged in order for visual extinction to occur.
- One of the main problems with this conclusion is that lesions following a stroke which damage the posterior parietal cortex are usually quite large, and could therefore affect other areas such as the occipital lobe, which could cause extinction to take place.
Meister et al (2006):
Study investigated the role of temporal and parietal areas in spatial attention. They gave healthy participants TMS to each areas while the participants had to identify unilateral or bilateral dots. They found that TMS to the TPJ caused extinction-like symptoms in healthy participants, whereas giving TMS to the superior temporal gyrus (STG) had no effect. Therefore, it can be concluded that posterior parietal areas are important in directing spatial attention to contralateral stimuli.
Ticini et al (2010):
This study also provided evidence that the TPJ is functioning incorrectly in patients with extinction, even if it is not lesioned. They carried out a perfusion-weighted MRI study using patients which had a damaged right basal ganglia – this imaging method facilitates measuring areas which are structurally intact but receive abnormal blood flow. They found that extinction patients with a lesion to the basal ganglia had a dysfunctional (but intact) right TPJ. The researchers concluded that damage to the basal ganglia could have affected projections from the putamen and caudate nucleus to the TPJ. This result shows both the importance of the TPJ to extinction, and also the role of white matter connections between various cortical and subcortical areas.
- Although these studies show the areas damaged in extinction, they do not show that it is distinct from neglect, as TMS can also induce mild neglect in healthy participants.
Chechlacz et al (2014):
Investigated the frequency and severity of extinction in the visual and tactile modalities in RH and LH strokes. 454 sub-acute stroke patients, with clinical notes and CT scans used to determine the type of stroke. Visual extinction task: had to identify if one or both fingers moved, in the tactile task, had to identify if right, left, or both hands touched. Results showed that both visual and tactile extinction were equally common after RH damage, while tactile was more common than visual after LH damage. Especially high frequency of extinction after RH damage affecting MCA & PCA areas. The severity of deficits did not vary.
- Results reflect RH dominance for disengaging attention (Corbetta & Shulman, 2002).
- Also agree with Kinsbourne (1987).
- The fact that severity did not differ could be because some patients had LH dominance – did not take handedness into account.
Karnath et al (2003):
They investigated the differences between areas involved in top-down and stimulus driven attention (the kind damaged in extinction). They analysed the extent of the lesions in patients who had visual extinction and spatial neglect using MRI, and found a dissociation between the areas involved in each disorder. Neglect patients showed a lesion overlap in the STG, while extinction patients had lesions which centred on the TPJ. Visual extinction patients also had lesions in the superior and middle temporal gyri and the ventral IPL, which suggests that these areas are also important for directing spatial attention.
- One of the strengths of this study is that they used patients who had only pure neglect syndrome or visual extinction – they had no visual field deficits caused by lesions in the occipital lobe.
- This is important as other cortical lesions could affect processing in the areas important for spatial attention.
Hillis et al (2006):
Argue that lesions to different cortical areas cause different types of extinction. Used MRI to evaluate the extent of patients’ lesions, and tested for visual, tactile, and motor extinction. Found that patients with visual extinction had damage to the visual association cortex, patients with tactile extinction had damage to the IPL, and patients with motor extinction had damage to the STG.
- More likely for patients to have visual and tactile extinction as the areas are nearby, so lesion could span both.
Vossel et al (2011):
They used voxel-based lesion-symptom mapping (VLSM) to analyse lesions of patients and found that participants with visual extinction had damage to the IPL, and that the greater the extent of the lesion, the more severe the extinction symptoms. In contrast, visual neglect was related to lesions in fronto-parietal and parietal-occipital cortex.
- This study also shows that visual neglect and extinction have a separate neural basis, although it identifies a different area of posterior parietal cortex to Karnath et al (2003).
Ҫiҫek et al (2007):
Study investigated the functional anatomy of unilateral and bilateral attention in an fMRI experiment using healthy participants. The participants were presented with stimuli on either the right or left, or bilaterally, and the stimuli were preceded by either valid or uninformative cue. They had to maintain fixation and respond as quickly as possible by pressing a button corresponding to the side the stimulus was presented on, if the stimulus was a target and not a distractor. They found activity in the IPL in the right hemisphere, the intraparietal sulcs (IPS) in both hemispheres, the inferior frontal gyrus in both hemispheres (IFG), the lateral peristriate region (LPS) in the right hemisphere, as well as the sensorimotor cortex and the cerebellum. They found that bilateral attention caused more activity in the dorsal parietal LPS and IFG than for unilateral attention.
- They therefore concluded that as the right IPS was activated more by bilateral stimulation, damage to this area could cause visual extinction.
Gilbert et al (2011):
They studied two patients who had lesions to the IPS: patient H.H. had lesion to the left IPS, and patient N.V. had a lesion to the right IPS. They had to carry out a Posner task where they had to respond to the area the stimuli were presented in. An invalid cue was presented before the stimulus on 17% of trials, and stimuli were presented bilaterally on 17% of trials. They also tested patients with IPL lesions to test whether they showed similar deficits in performance. The results showed that both patients had longer reaction times when the target was presented on the contralesional side with the presence of an ipsilateral distractor, and similar results were found with patients with IPL lesions. Patient N.V’s performance improved as their lesion got smaller, which suggests that the damage of the IPS is directionally proportional to extent of extinction. Further evidence for the importance of the IPS is shown by the fact that patient H.H. had intact IPL and TPJ. Therefore, this study seems to suggest that the IPS on either hemisphere is an important area for directing attention to the contralesional side of space.
- However, as similar results were found with the IPL patients it suggests that both areas are important in visual extinction.
Kinsbourne (1993):
States that a network of areas in the lesioned hemisphere causes deficits in directing attention to the contralesional side. Could explain why 3 different areas have been identified as being lesions in extinction: IPL, IPS and TPJ.


Unconscious Processing in Extinction
Several studies have shown that there is processing of extinguished items in patients with visual extinction.


Vuilleumier et al (2008):
Study examined the effects of unilateral parietal lobe damage on activation in the occipital cortex as a function of the ‘load’ of a task fixation. Patients had to make easy or hard judgements about a stimulus at fixation. While they were doing this, a checkerboard was flashed in LVF, RVF, or both. For bilateral presentation, there was reduced activity in the occipital lobe in the damaged hemisphere, particularly for the difficult task. This reduced activity could suggest why extinction occurs – limited attentional processing.
- Evidence for interactive processing between attention and perception.
Gilchrist et al (1996):
Presented participants with visual extinction with pairs of stimuli, and varied the different types of Gestalt factors between the stimuli. At baseline, participants reported seeing both items on 23% of trials. If the stimuli were the same colour, this went up to 60%, and collinearity was 83%. Gestalt factors aid perception, as they are processing in early visual regions, which are still intact.


Rees et al. (2000):
Researchers presented pictures of faces and houses to a patient with left extinction after a focal inferior parietal lesion. Using event-related fMRI to compare bilateral (where the left stimulus was extinguished) and unilateral right trials - which the patient perceived as the same - demonstrated that extinguished contralesional stimuli activated the striate cortex and early extrastriate visual areas in the damaged right hemisphere in a similar way to consciously perceived stimuli in the left hemisphere. The presentation of faces (compared to houses) also activated the fusiform face area, an area that may be specialised for processing faces (Kanwisher, McDermott & Chun, 1996).
Electrophysiological recordings of brain activity have also repeatedly been used to compare unconscious with conscious processing, with much the same findings as from imaging. For example, stimuli used by Rees et al. were presented to patients to look for the N170 (a visual evoked potential produced by faces) for extinguished faces. Compared to perceived faces, the contralesional stimuli induced similar early visual responses at electrodes over early occipital regions. Although these were slightly decreased in amplitude, the N170 was also found at posterior temporal electrodes. These results demonstrate that both early and higher-order visual areas can be activated unconsciously to some extent.
Mattingley et al (1997):
Case Study: Patient V.R. - stroke in the right middle cerebral artery which resulted in left extinction. Investigated whether V.R.’s extinction could be reduced when separate stimuli caused a perceptual ‘filling in’ to create a grouped stimulus. V.R. was presented with visual displays and had to indicate when any of the segments of circles had been removed from the display (exp. 1-3). Both inner and outer trials – inner is the Kanizsa shape. In experiment 4, she had to detect dots located within the Kanizsa figure. V.R responded verbally with ‘left, right, none, or both’ and accuracy was recorded. V.R showed much more extinction for bilateral displays for outer stimuli, as opposed to inner stimuli. This occurred even when the stimuli on the ipsilesional side were still facing in (Exp. 2) and in when the segment was removed but leaving the edge of the circle (Exp. 3). These act as control conditions. Exp. 4 was used to show that results were not caused by grouping of Kanizsa shapes, but that the space between them was filled in. V.R. again showed less extinction for Kanizsa shapes.
Also tested whether perceptual ‘filling in’ occurs with 3D shapes. V.R. had to identify whether the black bar was on both sides of the cube, with bilateral and unilateral trials. V.R. showed less extinction in the occlusion condition which suggests that she saw bar as one object.
- Evidence for unimpaired preattentive processing in extinction.
Di Pellegrino et al (2005):
Study measured extinction for objects which vary affordances – cups with handles affording an action for the left or right hand. Found that extinction was reduced when the handle of the cup was on the left, affording a left hand grasp. This occurred even though no motor response was required. Did not occur when the handles were replaced with a patch which covered the same visual area.
- Results suggest that action-related information is processed in extinction, even though there is no conscious awareness.
- Affordances enhance the strength of the object representation.
Wager et al (2014):
Study built on conclusions from Riddoch et al (2003) – that if objects are arranged for action then they are likely to form a single functional unit and are less likely to compete for representation. Tested this hypothesis using fMRI. Presented pps with pairs of semantically-related objects positioned correctly or incorrectly for action. Assessed competitive interactions by comparing signals produced when they were presented simultaneously or sequentially. Found that if they were presented correctly for action, they were like likely to compete for representation in V4.
- Results therefore suggest that a knowledge of objects and their interactions creates a larger ‘action unit’ that modulates early visual processing.