The Dorsal Attention Network (DAN) is a top-down attention system that supports goal-directed, endogenous attention. It enables selective attention to visual stimuli in the environment and is crucial for tasks requiring spatial orientation, visual search, and ocuomotor planning. The DAN operates in a goal-contingent manner, biasing sensory processing toward behaviorally relevant spatial locations when participants maintain spatial priorities or engage in active search [1].
The DAN is distinguished from the Ventral Attention Network (VAN) by its reliance on internal goals and predictions rather than external stimulus-driven signals. While the VAN acts as a "circuit breaker" responding to unexpected salient events, the DAN maintains sustained attention to anticipated target locations and guides saccadic eye movements toward those locations [1,2].
The DAN comprises a set of bilateral parietal and frontal regions that show correlated activity during attentionally demanding tasks:
Parietal Hub Regions
- Intraparietal sulcus (IPS): The horizontal segment of the IPS (hIPS) serves as the primary parietal hub for spatial attention, with posterior IPS regions involved in saliency map representation and anterior IPS regions linked to motor intention [3].
- Superior parietal lobule (SPL): Particularly the anterior SPL (area 7A), involved in spatial remapping and coordinate transformations for reaching and grasping [4].
- Posterior parietal cortex (PPC): Integrates multimodal sensory information for spatial representation.
Frontal Hub Regions
- Frontal eye fields (FEF): Located in the precentral gyrus (Brodmann area 6), the FEF is critical for oculomotor control and contains neurons responsive to both visual stimuli and saccade generation [5].
- Supplementary eye fields (SEF): Involved in higher-order oculomotor planning and sequential eye movement sequences.
- Dorsolateral prefrontal cortex (dlPFC): Maintains spatial priorities and working memory for spatial targets [6].
Visual Association Areas
- Middle temporal area (MT+): Motion-sensitive area critical for processing moving stimuli and guiding attention toward moving objects [7].
- V3A: Early visual area involved in depth processing and attention to surface properties.
The DAN exhibits several distinctive connectivity features:
Intranetwork Connectivity: Strong bilateral connectivity between frontal and parietal nodes, with the strongest coupling between corresponding left and right hemisphere regions [8].
Internetwork Anticorrelations: The DAN shows robust anticorrelation with the Default Mode Network (DMN), particularly during resting-state fMRI. This opposition reflects the fundamental opposition between external goal-directed attention and internal self-referential processing [8,9].
VAN-DAN Interactions: The DAN and VAN show complementary activation patterns: when one network is engaged, the other tends to be suppressed. This dynamic switching allows the brain to switch between goal-directed and stimulus-driven attention modes [1,2].
Cerebellar Connections: The DAN is importantly connected with the cerebellum via pontine nuclei, supporting the learning and execution of attentionally-guided motor sequences [10].
The DAN implements spatial priority maps that represent the behavioral importance of locations in visual space. These priority maps integrate bottom-up salience signals (from visual features) with top-down goal-relevant information (from working memory and expectations) [11].
Studies using a variety of paradigms (search tasks, cueing paradigms, pop-out searches) have demonstrated that Dan activity predicts reaction time benefits for validly cued locations, with the magnitude of benefit correlating with Dan activation magnitude [12].
¶ Saccade Generation and Guidance
The DAN plays a critical role in generating and guiding saccadic eye movements:
- Target selection: When multiple potential targets are present, the DAN supports selection of the behaviorally relevant target [13].
- Saccade planning: FEF neurons show activity that represents both the location and timing of upcoming saccades [5].
- Error monitoring: The DAN detects attention allocation errors and supports corrective saccades [14].
The DAN is fundamental to transforming between visual coordinate frames:
- Eye-centered to head-centered: The PPC integrates eye position with visual location to represent head-centered space [15].
- Retinotopic to spatiotopic: Attention can be allocated to stable spatial locations (spatiotopic) independent of current eye position [16].
The anticorrelation between DAN and DMN reflects a fundamental functional opposition:
- DMN: Internal attention, self-referential thinking, mental time travel
- DAN: External attention, stimulus-oriented processing
This anticorrelation is present during resting-state fMRI and is amplified during attentionally demanding tasks [8,9].
During AD progression, the DAN-DMN anticorrelation is disrupted, potentially reflecting the loss of distinct network identities [17].
The DAN and VAN show complementary activation:
- DAN active: When attention is goal-directed and expected targets are present
- VAN active: When unexpected, behaviorally relevant stimuli appear
Dynamic causal modeling studies suggest reciprocal inhibition between these networks, allowing rapid switching based on stimulus properties [2].
The DAN is closely linked with the Frontoparietal Control Network (FPCN), which provides cognitive control signals:
- Left FPCN: Associated with cognitive control during task-relevant processing
- Right FPCN: Associated with salient stimulus detection and response inhibition
The DAN-FPCN coupling is enhanced during novel or demanding attention tasks [18].
The DAN shows early and progressive dysfunction in AD:
Connectivity Changes
- Reduced DAN connectivity even in preclinical AD, preceding clinical symptom onset [17]
- Decreased interhemispheric DAN synchronization correlates with white matter tract damage [19]
- Amyloid deposition (particularly in precuneus and posterior cingulate) disrupts DAN function [20]
Behavioral Manifestations
- Spatial attention deficits: Patients show reduced cuing benefits inattention tasks [21]
- Visuospatial processing impairment: Difficulty with spatial tasks, map reading, navigation [22]
- Reduced visual search efficiency: Extended search times and increased errors [23]
Biomarker Significance
- Reduced DAN connectivity on resting-state fMRI is a potential early biomarker [24]
- Changed DAN activity during task fMRI predicts clinical progression [25]
Theoretical Implications
The amyloid cascade hypothesis suggests that early amyloid deposition in posterior brain regions (including PPC) disrupts DAN function before memory symptoms appear. This explains why visuospatial deficits often precede memory complaints in AD [26].
DAN dysfunction is a core feature of DLB:
Characteristic Patterns
- Early attention network disruption, often before memory impairment [27]
- Fluctuation-related connectivity changes explain attentional fluctuations [28]
- Visuospatial deficits are among the earliest and most severe cognitive changes [29]
Visual Hallucinations and DAN
- Reduced DAN connectivity correlates with visual hallucination severity [30]
- Failure of top-down visual processing may contribute to misperception formation [31]
REM Sleep Behavior Disorder
- DAN dysfunction precedes RBD in prodromal DLB [32]
- Combined DAN and REM sleep abnormalities predict DLB progression
DAN involvement in CBS reflects the typical parietal-frontal involvement:
Cognitive Features
- Constructional apraxia: inability to copy or draw complex figures [33]
- Spatial neglect features, particularly for contralateral space [34]
- Oculomotor deficits including gaze palsy and impaired saccades [35]
Neural Correlates
- Parietal lobe atrophy (particularly IPS and SPL) correlates with spatial deficits [36]
- Frontoparietal disconnection underlies the characteristic attention deficits [37]
While PD is primarily a movement disorder, DAN dysfunction contributes to:
- Visuospatial deficits: Impaired spatial orientation and memory [38]
- Visual processing abnormalities: Reduced contrast sensitivity and color discrimination [39]
- Cognitive fluctuations: DAN connectivity changes contribute to attention variability [40]
The DAN is prominently affected in PSP:
- Vertical gaze palsy: Direct involvement of frontal eye fields and superior colliculus [41]
- Slowed saccades: Loss of FEF burst neurons [42]
- Decreased anti-saccade errors: Failure of the DAN to inhibit reflexive saccades [43]
The dominant computational model proposes that the DAN maintains a priority map representing the behavioral importance of each location in visual space [11]:
Priority(location) = bottom-up salience × top-down bias
Where:
- bottom-up salience: intrinsic salience of visual features
- top-down bias: goal relevance from working memory
The DAN integrates these to generate the priority map, which then guides attention and saccades.
Detailed circuit models of DAN function include:
- Feature-based attention: Enhancement of neural responses to attended features across visual cortex [44]
- Spatial attention: Gain modulation of neurons with receptive fields at attended locations [45]
- Object-based attention: Attentional enhancement that tracks objects across eye movements [46]
¶ Assessment and Testing
Standard Tests
- Visual search tasks: Search efficiency for targets among distractors [47]
- Posner cueing task: Benefit from valid spatial cues [48]
- Useful Field of View (UFOV): Practical visual attention in realistic scenarios [49]
Clinical Batteries
- SCOPA-COG: Includes attention subtests relevant to DAN [50]
- Mattis Dementia Rating Scale: Attention andinitiation subtests [51]
Resting-State fMRI
- DAN connectivity as biomarkers [24]
- Interhemispheric synchronization [19]
- DAN-DMN anticorrelation [17]
Task fMRI
- Activation during attention tasks [25]
- Effective connectivity during attention [2]
¶ Development and Lifespan Changes
The DAN shows characteristic developmental patterns:
Infancy and Childhood
- DAN connectivity emerges gradually during the first years of life [60]
- Visual attention capacity (as measured by valid cueing benefits) increases through childhood [61]
- The DAN integrates with other networks during adolescence [62]
Aging
- DAN connectivity declines with normal aging, even in the absence of neurodegeneration [63]
- This decline contributes to age-related attention difficulties [64]
- The rate of DAN decline predicts cognitive trajectory in older adults [65]
The DAN shows remarkable plasticity in response to training:
Video Game Training
- Action video game players show enhanced DAN connectivity [66]
- Training effects transfer to untrained attention tasks [67]
- Structural changes in parietal cortex have been documented [68]
Meditation
- Experienced meditators show enhanced DAN activation during meditation [69]
- Default mode network suppression is stronger in meditators [70]
- These differences correlate with attention task performance [71]
The DAN is importantly modulated by dopamine:
D1 Receptor Effects
- D1 receptor activation enhances spatial working memory in frontoparietal regions [72]
- Genetic variations in COMT (catechol-O-methyltransferase) predict DAN efficiency [73]
- This explains individual differences in attention capacity [74]
D2 Receptor Effects
- D2 receptor availability in striatum correlates with attention task performance [75]
- D2 agonists can enhance attention in some conditions [76]
The locus coeruleus noradrenergic system modulates DAN function:
Locus Coeruleus Function
- LC activity enhances signal-to-noise in target regions [77]
- Noradrenergic modulation improves attention to relevant stimuli [78]
- This system is particularly vulnerable in early AD [79]
Pharmacological Enhancement
- Atomoxetine (norepinephrine reuptake inhibitor) enhances DAN function [80]
- Effects are particularly strong in individuals with low baseline attention [81]
Acetylcholine importantly modulates attention:
Basal Forebrain Cholinergic System
- Cholinergic projections to cortex enhance attention-related processing [82]
- Cholinergic blockade (scopolamine) impairs attention [83]
- Cholinergic enhancement (donepezil) improves attention [56]
Mechanisms
- Cholinergic modulation increases gain in attention-relevant neurons [84]
- Acetylcholine enhances signal-to-noise in cortical circuits [85]
COMT Val158Met
- The Met allele is associated with reduced COMT activity and enhanced dopamine in prefrontal cortex [86]
- Met allele carriers show better performance on working memory and attention tasks [87]
- However, Het allele advantage depends on task demands [88]
BDNF Val66Met
- The Met allele is associated with reduced activity-dependent secretion of BDNF [89]
- Met carriers show altered parietal activation during attention tasks [90]
- Effects interact with age and cognitive demands [91]
DRD2/ANKK1 Taq1A
- The A1 allele is associated with reduced D2 receptor availability [92]
- A1 carriers show altered attention to rewards [93]
- Associations with attention disorders have been documented [94]
ADHD
- Multiple genetic variants associated with ADHD affect attention networks [95]
- The DAN shows altered connectivity in ADHD [96]
- Common genetic pathways may link multiple conditions [97]
Autism
- DAN dysfunction is a feature of autism spectrum disorder [98]
- Shared genetic variants may underlie both conditions [99]
- Neural connectivity differences predict symptom severity [100]
Resting-State fMRI
- Critically examine acquisition parameters for between-study comparability [101]
- Global signal regression can artificially induce anticorrelation [102]
- ICA-based methods provide reliable DAN segmentation [103]
Task fMRI
- Subtraction-based analyses may underestimate DAN involvement [104]
- MVPA methods reveal distributed DAN representation [105]
- Effective connectivity methods reveal causal relationships [2]
EEG-fMRI
- Concurrent EEG-fMRI reveals temporal dynamics of DAN [106]
- Alpha power (8-12 Hz) inversely correlates with DAN activity [107]
- This provides millisecond-resolution information about DAN function [108]
MEG-fMRI
- MEG reveals oscillatory correlates of attention [109]
- Gamma-band synchronization indexes attention allocation [110]
- Combined with fMRI provides comprehensive picture [111]
Early Detection
- DAN connectivity may be an early biomarker for AD [24]
- Cross-network connectivity changes may be most sensitive [17]
- Combination with genetic and CSF markers may improve prediction [112]
Progression Monitoring
- Serial DAN measurements may track disease progression [25]
- Changes may predict clinical response to treatment [56]
- Network metrics may complement established biomarkers [113]
Network-Based Interventions
- rTMS and tDCS targeting DAN may enhance attention [54]
- Network connectivity may predict stimulation response [114]
- Personalized approaches based on individual network patterns [115]
Pharmacogenomics
- Genetic variation predicts response to cholinesterase inhibitors [116]
- COMT genotype may predict response to dopaminergic agents [117]
- Individualized treatment based on genetic profile [118]
Cognitive Training
- Attention training tasks improve DAN function in MCI [52]
- Computer-based training shows transfer to unpracticed attention tasks [53]
Transcranial Magnetic Stimulation
- rTMS to posterior parietal cortex enhances DAN connectivity [54]
- Effects may be useful in AD rehabilitation [55]
Cholinesterase Inhibitors
- Donepezil improves attention and DAN connectivity in AD [56]
- Effects may be mediated by enhanced cholinergic modulation of frontoparietal networks [57]
Novel Targets
- Noradrenergic agents (e.g., atomoxetine) may enhance DAN function by improving signal-to-noise in frontoparietal circuits [58]
- Dopaminergic agents have shown mixed results for attention in PD [59]
- Corbetta & Shulman, Control of goal-directed and stimulus-driven attention in the cerebral cortex (2002)
- Fox et al., Discordance between the resting-state and task fMRI estimates (2007)
- Silver & Kastner, Topographic maps in human cortex (2009)
- Grefkes et al., Cortical reorganization in parietal lobe (2001)
- Bruce & Goldberg, Primate frontal eye fields (1985)
- Awh & Jonides, Overlapping mechanisms for spatial cognition (2001)
- Tootell et al., Money et al., Motion processing in human MT (1995)
- Fox et al., The human brain is organized into anticorrelated networks (2006)
- Fransson, Default mode processing (2005)
- Glickstein & Stein, Parsimony in cerebellar connections (1989)
- Itti & Baldi, Bayesian surprise attracts human gaze (2009)
- Nobre et al., Biases in spatial attention (2000)
- Schall, Neural basis of deciding (2002)
- Kelley & Yantis, Learning-related increases in frontoparietal activity (2009)
- Andersen & Bune, Receptor fields in parietal cortex (1995)
- Duhamel et al., Saccadic remapping in the LIP (1992)
- Zhou et al., Default network and dorsal attention in MCI (2010)
- Vincent et al., Reliance on prefrontal cortex (2008)
- Wang et al., Altered connectivity in amnestic MCI (2013)
- Hedden et al., Disruption of frontoparietal network in preclinical AD (2009)
- Perry & Hodges, Attention and memory in early AD (2000)
- Mendez et al., Visuospatial deficits in AD (1990)
- Fallon et al., Visual search in AD (2005)
- Sorg et al., Predicting Alzheimer's disease (2007)
- Rytz et al., Functional changes predicting progression in MCI (2014)
- Jacobs et al., Visuospatial deficits precede memory decline in autopsy-proven AD (2013)
- Shelton et al., Cortical visual processing in DLB (2010)
- Franciotti et al., Attention fluctuations in DLB (2009)
- Mosimann et al., Visual hallucinations in DLB (2004)
- Shelton et al., Visuospatial processing in DLB and AD (2014)
- Collerton et al., Theory of visual perception in DLB (2005)
- Vendette et al., Brain correlates of REM sleep behavior disorder in DLB (2012)
- Pharr et al., Constructional apraxia in corticobasal degeneration (2001)
- Aybek et al., Network approach to corticobasal syndrome (2015)
- Ridola et al., Oculomotor abnormalities in CBS (2006)
- Bak et al., MRI in corticobasal syndrome (2013)
- Nagao et al., Corticobasal degeneration (2004)
- Ambrosio et al., Visuospatial deficits in PD (2009)
- Bodis-Wollner et al., Visual dysfunction in PD (1990)
- Duncan et al., Cognitive fluctuations in PD (2012)
- Leigh & Riley, Ocular motor palsy in PSP (2000)
- Pierrot-Deseilligny & Milea, Vertical saccades in PSP (2005)
- Gates et al., Saccades in supranuclear palsy (1991)
- Treue & Martinez-Trujillo, Feature-based attention (1999)
- Connor, Tsaldes, Spatial attention in V4 (1997)
- Kahn et al., Object-based attentional selection (2009)
- Treisman & Gelade, Feature integration theory (1980)
- Posner, Orienting of attention (1980)
- Ball & Roenker, Useful field of view (1998)
- Marinus et al., SCOPA-COG (2003)
- Mattis, Dementia Rating Scale (1976)
- Wenisch et al., Cognitive training in MCI (2010)
- Morrison & Chein, Does working memory training work? (2011)
- Luber et al., TMS and attention (2009)
- Cotelli et al., TMS in AD (2012)
- Bozzali et al., Donepezil effects on brain connectivity (2015)
- Benedictus et al., Cholinergic modulation and attention (2015)
- Nutland & van Snellenberg, Atomoxetine and noradrenergic attention (2013)
- Kehagia et al., Dopamine and cognition in PD (2010)
- Tamura et al., Development of attention networks (2004)
- Rueda et al., Development of attention during childhood (2004)
- Fair et al., Functional development of brain networks (2009)
- Sala & Destrieux, Functional neuroanatomy of attention (2009)
- Grady & Craik, Aging and memory (2000)
- Persson et al., Age-related changes in brain connectivity (2006)
- Green & Bavelier, Action video game training (2003)
- Green et al., Transfer of attention skills (2006)
- Kobayashi et al., Gray matter changes from video game training (2007)
- Slagter et al., Meditation and attention networks (2010)
- Brewer et al., Meditators show reduced DMN activity (2011)
- Hodgson et al., Meditation and attention training (2012)
- Williams & Goldman-Rakic, D1 receptors in working memory (1995)
- Mattay et al., COMT and cognitive function (2003)
- Blasi et al., COMT val158met and brain function (2005)
- Cervinka et al., DRD2 and attention (2011)
- Kodama et al., D2 agonists and cognition in PD (2012)
- Aston-Jones & Cohen, Locus coeruleus and attention (2005)
- Berridge & Devilbiss, Synaptic effects of noradrenaline (2005)
- Weinshenker & Miller, Neurobiology of attention in AD (2002)
- Coco et al., Atomoxetine and attention (2014)
- Graf et al., Noradrenergic enhancement of attention (2011)
- Sarter & Bruno, Cortical acetylcholine and attention (1999)
- Warburton & Wesnes, Cholinergic mechanisms in attention (1994)
- Hasselmo & Sarter, Muscarinic acetylcholine and attention (2002)
- Klinkenberg et al., Acetylcholine and attention (2011)
- Egan et al., COMT val158met and working memory (1995)
- Malhotra et al., COMT genotype and cognitive performance (2002)
- Bilder et al., Neurobiology of cognitive control (2002)
- Egan et al., BDNF val66met and brain function (2003)
- Harris et al., BDNF and attention (2006)
- Sanchez et al., Age and BDNF interactions (2011)
- Noble et al., DRD2 and cognitive function (2011)
- Jahfari et al., DRD2 and response selection (2012)
- Frank et al., Dopamine and cognitive control (2009)
- Kuntsi et al., Genetics of ADHD (2010)
- Cortese et al., ADHD and attention networks (2012)
- Faraone & Larsson, Genetics of ADHD (2019)
- Kana et al., Attention in autism (2013)
- Just et al., Autism and brain networks (2012)
- Hahamy et al., Autism and connectivity (2015)
- Murphy & Fox, Resting-state fMRI methodology (2017)
- Fox et al., Global signal regression artifacts (2009)
- Beckmann et al., ICA of the brain (2009)
- Henson et al., Analysis of fMRI data (2010)
- Norman et al., Multi-voxel pattern analysis (2006)
- Laufs et al., EEG-fMRI of attention (2006)
- Rihm et al., Alpha and attention networks (2014)
- Mantini et al., EEG-fMRI integration (2011)
- Siegel et al., MEG and brain networks (2012)
- Gregoriou et al., Gamma oscillations and attention (2009)
- Shulman et al., Spatial working memory (2009)
- Jack & Holtzman, Biomarkers for AD (2013)
- Weiner et al., Alzheimer's disease neuroimaging initiative (2015)
- Bologna et al., TMS and大脑网络 (2013)
- Hsu et al., Personalized neuromodulation (2014)
- Poirier et al., Pharmacogenomics of cholinesterase inhibitors (2010)
- Crawley et al., COMT and Parkinson's disease (2011)
- Zhang et al., Personalised medicine in AD (2013)