Special issue: Research reportHuman dorsolateral prefrontal cortex is involved in visual search for conjunctions but not features: A theta TMS study
Introduction
Detecting and locating a target in the presence of distractors constitutes a basic function of the visual system. Despite its apparent straightforward nature, a number of brain areas have been implicated in performing the task, particularly when the object being searched for is identified by a combination of attributes, and activates regions in the visual, parietal and frontal cortices (Donner et al., 2000). The necessity of the involvement of some of the areas implicated by both lesion findings (e.g., Collin et al., 1982, Eglin et al., 1991) and imaging studies in humans (e.g., Donner et al., 2002; Makino et al., 2004) has been tested using transcranial magnetic stimulation (TMS) (Ashbridge et al., 1997, Walsh et al., 1998a, Muggleton et al., 2003). This technique has shown that both frontal eye fields (FEF) and posterior parietal cortex (PPC) regions are necessary for detection of a search target defined by a conjunction of features, but not for visual search tasks where a single quality of the stimulus, such as colour, defines the target (Ashbridge et al., 1997, Walsh et al., 1998a, Ellison and Walsh, 1998). Furthermore, the timing of the contribution of these two areas has been dissociated, with an earlier involvement for FEF than PPC (O'Shea et al., 2004, Kalla et al., 2008: see also Juan et al., 2008).
The role of the dorsolateral prefrontal cortex (DLPFC) in visual search has yet to be probed with TMS. This area is frequently associated with working memory and this is consistent with a hypothesised role for memory in visual search (Wolfe, 1994, Treisman and Gelade, 1980) in the prevention of revisiting already searched locations. In line with this view, there is evidence of a bias away from recently examined locations e.g., Klein and McInnes (1999). However, use of a paradigm in which the locations of the array elements was changed frequently during a trial did not reveal any deleterious effects on search performance (Horowitz and Wolfe, 1998). The absence of a relationship between working memory capacity and search set-size/reaction time slopes (Kane et al., 2006) may be more consistent with memory being important in representing the target, rather than remembering the searched locations. Despite disagreements on the role played by memory in search, there is broad, but not complete, agreement that it does indeed play a role (Beck et al., 2006, Kristjansson, 2000). If this were the case then DLPFC would be a prime candidate for fulfilling this role. Common activations in DLPFC have been reported in the same subjects performing visual search and memory search (Makino et al., 2004). They proposed that the common process was the monitoring and manipulating of multiple elements and that target matching was another possible function of this area, a cognitive process previously associated with both visual and memory search (Hillstrom and Logan, 1998).
We tested the hypothesis that DLPFC is, like FEF and PPC, necessary for efficient performance of conjunction visual search. In doing so, we employed methods similar to those used in our previous investigations of these other areas (Muggleton et al., 2008, Muggleton et al., 2003, O'Shea et al., 2004, O'Shea et al., 2007). The visual area MT/V5, typically involved in motion processing was selected as a control site for non-specific visual system effects as it is involved in visual processing, but was not expected to be involved in colour-form conjunction search. A signal detection theory (SDT) approach was used, with performance measured by means of the sensitivity measure d′, as well a bias measure which allowed the tendency to miss the target or make false alarms to be quantified. Due to the potential discomfort that might be expected to result from TMS delivered over DLPFC, a consequence of its relatively anterior location resulting in less tolerable effects on muscles/nerves in the vicinity, we employed theta TMS (Huang et al., 2005, Vallesi et al., 2007). Stimulation applied in this manner results in prolonged disruption of the targeted area as a consequence of a relatively short period of TMS delivery (seconds rather than the multiple minutes required for the more typical offline stimulation used, in which TMS is delivered at 1 Hz). As testing takes place in the post-TMS period, any interference with performance attributable to the effects of discomfort or blinks and facial twitches is avoided.
Section snippets
Subjects
Twelve subjects (8 males, 4 females, mean age 28.3 years) were tested on each experimental condition. The same subjects performed all three conjunction conditions. Seven of them also performed the feature search task, along with five new subjects (6 males, 6 females, mean age 27.5 years). For six subjects in the conjunction task this experiment was their first experience of TMS and the same was the case for five subjects for the feature search task. All subjects gave informed consent prior to
Results
Data were analysed by means of d′ scores calculated for each time period and condition in the experiment. d′ Scores were calculated as the difference between the z scores of the proportion of correct responses on the target present trials (hit rate, h) and the proportion of incorrect responses on target-absent trials (false alarm rate, f) according to the equation: d′ = z(h)−z(f) (where the z scores represent the area under a normally distributed curve with a mean of 0 and a standard deviation of
Conclusions
These findings confirm the prediction that DLPFC is necessary for successful visual search for a target defined by a conjunction of features but not one defined by a single attribute. Theta burst TMS delivered over this area caused a significant reduction in d′ only for the former task and was without effect on the latter. As expected, vertex TMS was without effect on the conjunction task. In contrast, TMS delivered over MT/V5 (the control site) resulted in improved conjunction search
References (40)
- et al.
Temporal aspects of visual search studied by transcranial magnetic stimulation
Neuropsychologia
(1997) - et al.
The role of frontal eye-fields and superior colliculi in visual search and non-visual search in rhesus monkeys
Behavioral Brain Research
(1982) - et al.
Visual feature and conjunction searches of equal difficulty engage only partially overlapping frontoparietal networks
Neuroimage
(2002) - et al.
Perceptual learning in visual search: some evidence of specificities
Vision Research
(1998) - et al.
Theta burst stimulation of the human motor cortex
Neuron
(2005) - et al.
Bayesian adaptive estimation of psychometric slope and threshold
Vision Research
(1999) - et al.
Visual search and memory search engage extensive overlapping cerebral cortices: an fMRI study
Neuroimage
(2004) - et al.
The role of the angular gyrus in visual conjunction search investigated using signal detection analysis and transcranial magnetic stimulation
Neuropsychologia
(2008) - et al.
Motor and phosphene thresholds: a transcranial magnetic stimulation correlation study
Neuropsychologia
(2001) - et al.
A feature-integration theory of attention
Cognitive Psychology
(1980)
Cortical plasticity in perceptual learning demonstrated by transcranial magnetic stimulation
Neuropsychologia
Memory for where, but not what, is used during visual search
Journal of Experimental Psychology: Human Perception and Performance
Visual area V5/MT remembers “what” but not “where”
Cerebral Cortex
Priming of motion direction and area V5/ MT: a test of perceptual memory
Cerebral Cortex
Involvement of the human frontal eye field and multiple parietal areas in covert visual selection during conjunction search
European Journal of Neuroscience
A new anatomical landmark for reliable identification of human area V5/MT: a quantitative analysis of sulcal patterning
Cerebral Cortex
Cortical substrates supporting visual search in humans
Cerebral Cortex
Distributed hierarchical processing in the primate cerebral cortex
Cerebral Cortex
Enhanced visual spatial attention ipsilateral to rTMS-induced 'virtual lesions' of human parietal cortex
Nature Neuroscience
Decomposing visual search: evidence of multiple item-specific skills
Journal of Experimental Psychology: Human Perception and Performance
Cited by (42)
Evaluating the causal contribution of fronto-parietal cortices to the control of the bottom-up and top-down visual attention using fMRI-guided TMS
2020, CortexCitation Excerpt :This question remains to be clarified in future studies with appropriate task manipulations and methodology, such as combining TMS with event-related potentials or eye movements. Furthermore, in previous studies the selection of the TMS stimulation sites was either based on other related TMS/fMRI studies (Kalla et al., 2009; Yan et al., 2016) or on anatomical descriptions (Heinen, Feredoes, Weiskopf, Ruff, & Driver, 2014; Muggleton et al., 2003, pp. 3340–3343). In the present study, we used subject-specific ROIs obtained from the fMRI scan, making the simulation site more precise, as well as using post-hoc fMRI analysis to corroborate the TMS findings.
Prefrontal cortex and impulsivity: Interest of noninvasive brain stimulation
2016, Neuroscience and Biobehavioral ReviewsCitation Excerpt :Studies with rTMS showed that both PFC hemispheres were involved in the modulation of sustained attention (Table 1). High frequency rTMS applied on the left or right PFC resulted in increased sustained attention (Vanderhasselt et al., 2006, 2007; Hwang et al., 2010) whereas cTBS stimulation reduced it (Kalla et al., 2009). Such modulation of sustained attention may only be effective in higher level cognitive tasks, since Vanderhasselt et al. (2007) showed that high‐frequency rTMS applied on the right PFC reduced subjects’ reaction time to congruent and incongruent stimuli in high‐level cognitive tasks (Stroop task with 80% of incongruent stimuli (i.e., word and color of the word are different such as the word “blue” written with red ink)).
'How many' and 'how much' dissociate in the parietal lobe
2015, CortexCitation Excerpt :This has been taken as evidence that numerosity estimation (how many) is independent from other continuous dimensions (how much) (e.g., Halberda & Feigenson, 2008; Halberda et al., 2008; Piazza et al., 2010). Some more recent studies have, however, challenged this view suggesting that instead numerosity processing is intrinsically dependent from or modulated by continuous quantity (Cappelletti, Didino, Stoianov, & Zorzi, 2014, 2015; Dakin, Tibber, Greenwood, Kingdom, & Morgan, 2012; Gebuis & Gevers, 2011; Gebuis & Reynvoet, 2012a, 2012b, 2014; Henik, Leibovich, Naparstek, Diesendruck, & Rubinsten, 2012; Hurewitz, Gelman, & Schnitzer, 2006; Mix, Huttenlocher, & Levine, 2002; Szucs, Devin, Soltesz, Nobes, & Gabriel, 2013). These studies are based on the distinction between ‘congruent’ and ‘incongruent’ numerosity trials, whereby numerosity either correlates with or instead is inversely correlated to continuous variables (Cappelletti et al., 2014, 2015; Gebuis & Gevers, 2011; Gebuis & Reynvoet, 2012a, 2012b, 2014; Szucs et al., 2013).
Timing of spatial priming within the fronto-parietal attention network: A TMS study
2015, Neuropsychologia