Abstract

There is emerging evidence that alterations in dopaminergic transmission can influence semantic processing, yet the neural mechanisms involved are unknown. The influence of levodopa (L-DOPA) on semantic priming was investigated in healthy individuals (n = 20) using event-related functional magnetic resonance imaging with a randomized, double-blind crossover design. Critical prime–target pairs consisted of a lexical ambiguity prime and 1) a target related to the dominant meaning of the prime (e.g., bank–money), 2) a target related to the subordinate meaning (e.g., fence–sword), or 3) an unrelated target (e.g., ball–desk). Behavioral data showed that both dominant and subordinate meanings were primed on placebo. In contrast, there was preserved priming of dominant meanings and no significant priming of subordinate meanings on L-DOPA, the latter associated with decreased anterior cingulate and dorsal prefrontal cortex activity. Dominant meaning activation on L-DOPA was associated with increased activity in the left rolandic operculum and left middle temporal gyrus. These findings suggest that L-DOPA enhances frequency-based semantic focus via prefrontal and temporal modulation of automatic semantic priming and through engagement of anterior cingulate mechanisms supporting attentional/controlled priming.

Introduction

The role of dopamine (DA) in human cognition, including its possible neural mechanisms, has been primarily investigated with respect to working memory and learning (e.g., Cools 2006; Cools et al. 2007; Rowe et al. 2008). Yet, there is also emerging evidence that DA modulation influences language function and, in particular, semantic processing in healthy subjects (Kischka et al. 1996; Copland, Chenery, et al. 2003; Angwin et al. 2004; Roesch-Ely et al. 2006) and in neurocognitive disorders such as Parkinson's disease (PD) (Angwin, Chenery, et al. 2006) and schizophrenia (Goldberg et al. 2000). However, the nature of the neuromodulatory role played by DA in semantic processing is unclear, as are the neural systems it engages.

The present study employed a semantic priming functional magnetic resonance imaging (fMRI) task to investigate the neuromodulatory role of DA in semantics. The semantic priming effect refers to the increased speed and accuracy in word recognition tasks, such as lexical decisions (word/nonword), made on target words when preceded by a semantically or associatively related prime word (e.g., doctor–nurse), compared with an unrelated word (e.g., bread–nurse). Semantic priming effects are assumed to reflect automatic spreading activation within semantic memory and/or attentional processes including expectancy generation, where the subject develops an expectancy set based on the prime, and postlexical semantic checking, where the relationship between the target and the prime is judged (Neely 1991). It is generally assumed that employing a short prime–target interval or stimulus-onset asynchrony (SOA) taps into primarily automatic priming mechanisms, whereas controlled or attentional priming mechanisms are invoked with longer SOAs and by increasing expectancies by manipulations of instructions or the proportion of related targets (Neely 1991).

Kischka et al. (1996) first observed that, compared with a placebo group, healthy individuals who have ingested L-DOPA (a DA precursor) show reduced indirect priming (e.g., summer–snow). This observation led to the proposal that DA focuses spreading activation within semantic networks, reducing the activation of distantly related items, consistent with increasing signal-to-noise ratio (SNR) of information processing. A subsequent study found that L-DOPA ingestion by healthy subjects performing a masked semantic priming task appeared to both speed up semantic activation at a short SOA and attenuate both direct and indirect priming effects at longer SOAs (Angwin et al. 2004), consistent with a role in accelerating lexical activation in terms of both gain and decay but not fully in keeping with an enhanced SNR within semantic networks as proposed previously (Kischka et al. 1996). A more recent study observed a trend toward decreased indirect priming for targets presented to the right visual field/left hemisphere when healthy subjects received pergolide (a D1/D2 agonist) but not bromocriptine (a D2 agonist) (Roesch-Ely et al. 2006), suggesting that DAergic semantic modulation may occur via D1 mechanisms influencing semantic associations. Although the use of a divided visual field paradigm indirectly implicates left hemisphere involvement in DA neuromodulation of semantics, the actual neural substrates remain unknown.

Lexical ambiguities (words with multiple meanings) provide an ideal avenue for investigating how DA influences the processing of competing representations (Cohen and Servan-Schreiber 1992). In the absence of context, meanings of ambiguities are selected and suppressed on the basis of preexisting meaning biases (Simpson and Burgess 1985), thus providing a means of testing the ability to select and inhibit competing information in the domain of semantic processing. The lexical ambiguity priming task employed in the present study involves priming of dominant-related (e.g., bank–money) versus subordinate-related (e.g., bank–river) word pairs. This paradigm is sensitive to semantic processing alterations in PD (Copland 2003) and to L-DOPA–induced priming modulation in healthy individuals (Copland, Chenery, et al. 2003), where reduced subordinate meaning priming was observed in healthy individuals who have ingested L-DOPA (Copland, Chenery, et al. 2003), providing further evidence of increased focusing of activation within semantic networks.

To examine the neuromodulation of semantic processing by DA, we investigated the brain mechanisms associated with L-DOPA–induced modulation of semantic priming in healthy individuals with fMRI. L-DOPA is taken up into DAergic neurons, enzymatically converted to DA, and can be released into the synaptic cleft in a phasic manner (Robinson et al. 2005). In the present study of lexical ambiguity priming on L-DOPA, we employed a short prime–target SOA of 150 ms to capture the activation of multiple competing meanings (Simpson and Burgess 1985). Cohen and Servan-Schreiber (1992) have argued that the selection of meanings for lexical ambiguities represents a situation where a change in DAergic activity may alter behavior, as witnessed in schizophrenia. As per our previous behavioral findings, we predicted that ingestion of L-DOPA in healthy individuals would enhance dominant meaning activation and/or reduce subordinate meaning activation, as opposed to the comparable priming of dominant and subordinate representations on placebo (Copland, Chenery, et al. 2003).

Previous neuroimaging studies of semantic priming indicate that both the regions engaged, and the direction of blood oxygen level–dependent (BOLD) signal change for related versus unrelated conditions varies as a function of the degree of automatic versus controlled processing mechanisms engaged (typically via manipulations of the prime–target interval) and the type of relationship between the prime and target (e.g., Tivarus et al. 2006). Previous event-related neuroimaging studies of semantic priming that have employed a lexical decision task with a short prime–target interval (see Table 1) suggest that more automatic aspects of semantic priming typically engage portions of the left temporal lobe (including superior, middle, and inferior temporal regions) (Kotz et al. 2002; Copland, de Zubicaray, et al. 2003; Rissman et al. 2003; Gold et al. 2006) and frontal regions including the left inferior frontal gyrus (Kotz et al. 2002; Copland, de Zubicaray, et al. 2003) and middle frontal gyrus (MFG, Kotz et al. 2002; Rissman et al. 2003). Controlled or attentional forms of semantic priming have been associated with modulated activity in the left middle temporal gyrus, left inferior prefrontal cortex, and the anterior and posterior cingulate (Matsumoto et al. 2005; Gold et al. 2006; Tivarus et al. 2006; Copland et al. 2007; O'Hare et al. 2008).

Table 1

Main regions involved in automatic semantic priming observed in event-related lexical decision fMRI studies

 Middle/superior temporal Inferior temporal Inferior frontal Middle frontal 
Copland, de Zubicaray, et al. 2003 L MTG  L IFG  
Gold et al. 2006  L fusiform   
Kotz et al. 2002 L MTG, planum temporale  L IFG, frontal operculum L MFG, R MFG 
Rissman et al. 2003 L STG   L MFG, R MFG 
 Middle/superior temporal Inferior temporal Inferior frontal Middle frontal 
Copland, de Zubicaray, et al. 2003 L MTG  L IFG  
Gold et al. 2006  L fusiform   
Kotz et al. 2002 L MTG, planum temporale  L IFG, frontal operculum L MFG, R MFG 
Rissman et al. 2003 L STG   L MFG, R MFG 

Note: L, left; MTG, middle temporal gyrus; IFG, inferior frontal gyrus; R, right; STG, superior temporal gyrus.

As the present study employed a short prime–target interval, it was hypothesized that L-DOPA would modulate activity within the network of regions engaged during automatic semantic priming discussed above. Of these regions, the prefrontal cortex, and particularly the dorsolateral prefrontal cortex, has been strongly implicated in DAergic modulation of cognition (Arnsten 1998; Mattay et al. 2002; Cools 2006). An alternative hypothesis is that DA enhances attentional capacity during information processing (Clark et al. 1987), and subjects on L-DOPA will therefore engage a network of regions associated with more attentional or strategic forms of semantic priming. Of these regions, the anterior cingulate has been linked to DAergic modulation of cognition in both healthy individuals and in PD (Mattay et al. 2002; Aalto et al. 2005). Given the proposal that DA may enhance semantic salience and/or accelerate lexical processing, the behavioral effects we expected were an increased selectivity of meaning activation either through increased priming of dominant targets and/or inhibition of subordinate targets on L-DOPA relative to placebo.

Materials and Methods

Subjects

Twenty young healthy right-handed volunteers (8 females) were recruited from The University of Queensland (mean age = 27.3 years; mean education = 15.3 years). Exclusion criteria included any psychotropic, antidepressant, or antihypertensive medications; pregnancy; and any history of mental illness or neurological disease or trauma. Subjects undertook a medical interview to confirm suitability for the study. This study was approved by The University of Queensland Medical Research Ethics Committee. All subjects gave informed written consent prior to participation and received $30 per session for participation.

Drugs and Study Design

A randomized double-blind, placebo-controlled crossover design was employed. Subjects were randomly assigned to receive either placebo or L-DOPA in session 1 and the other treatment in session 2 at least 6 days later. Order of drug and placebo was randomized and administered by a research nurse (blind to participants and investigators). A Madopar capsule containing 100 mg L-DOPA and 25 mg benserazied (a peripheral decarboxylase blocker) was administered. Testing took place 45 min after ingestion of the capsule. This time delay is consistent with peak plasma L-DOPA levels (Olanow et al. 2000). The dose is identical to that previously used in semantic priming tasks (Kischka et al. 1996), and we have used this dosage and time delay successfully in a cohort of 39 university students to demonstrate altered semantic priming without any adverse side effects (Angwin, Chenery, et al. 2006). Blood pressure and heart rate were taken prior to tablet ingestion and following the scan. Plasma DA levels were measured (blood samples were taken immediately prior to scanning) by reverse-phase high-performance liquid chromatography with electrochemical detection. The mood of subjects was measured immediately prior to scanning using the Bond and Lader measure of subjective feelings (Bond and Lader 1974).

Priming Experiment and Procedure

The priming experiment was adapted from a previous fMRI study of lexical ambiguity priming in healthy individuals (Copland, de Zubicaray, et al. 2003). Lexical ambiguities (words with one form and multiple independent meanings) were selected from norms (Nelson et al. 1980; Twilley et al. 1994) and further pretested on a separate group of young healthy participants to ensure that all ambiguities had 2 independent meanings, a dominant meaning provided at least 70% of the time by respondents and a subordinate meaning provided less than 30% of the time. One hundred and fifty lexical ambiguities were used as primes paired with targets that were a dominant associate (e.g., bank–money), a subordinate associate (e.g., fence–sword), an unrelated associate (e.g., ball–desk), or a legal nonword (e.g., mint–blark). Each critical condition consisted of 30 word pairs. There were 60 word–nonword pairs. There was no significant difference (P > 0.05) between the primes of the critical conditions in terms of word length, frequency (Kucera and Francis 1967), and concreteness (MRC Database) (see Table 2 for stimuli characteristics). There was no significant difference (P > 0.05) between the matched targets of the different conditions in terms of length, frequency, concreteness, or lexical decision reaction time in isolation (Balota et al. 1999).

Table 2

Experimental stimuli characteristics

 Word length (letters) Frequencya Concreteness score (100–700) 
Dominant primes 4.2 (0.8) 57.4 (52.3) 519.8 (91.1) 
Subordinate primes 4.5 (0.8) 39.4 (33.7) 530.7 (75.2) 
Unrelated primes 4.8 (1.1) 54.1 (67.3) 470.5 (103.1) 
Dominant targets 4.2 (0.8) 85.7 (52.8) 534.7 (98.9) 
Subordinate targets 4.3 (0.8) 79.6 (79.2) 499.9 (99.5) 
Unrelated targets 4.4 (0.7) 85.6 (103.2) 449.1 (105.8) 
 Word length (letters) Frequencya Concreteness score (100–700) 
Dominant primes 4.2 (0.8) 57.4 (52.3) 519.8 (91.1) 
Subordinate primes 4.5 (0.8) 39.4 (33.7) 530.7 (75.2) 
Unrelated primes 4.8 (1.1) 54.1 (67.3) 470.5 (103.1) 
Dominant targets 4.2 (0.8) 85.7 (52.8) 534.7 (98.9) 
Subordinate targets 4.3 (0.8) 79.6 (79.2) 499.9 (99.5) 
Unrelated targets 4.4 (0.7) 85.6 (103.2) 449.1 (105.8) 
a

In words per million (Kucera and Francis 1967).

Following a warning, the prime word was presented for 150 ms followed by target presentation for 2 s, during which time the response was recorded. Five trial types (4 priming conditions in addition to approximately 120 fixation [null] trials) were presented randomly, across 2 imaging sessions; 75 word pairs were presented in each session. The intertrial intervals were randomized to ensure optimal statistical efficiency (Dale 1999; range = 5.85–9.53 s, mean = 7.69 s). For the experiment, stimuli were enlarged and back projected onto the center of a screen at the foot of the bore of the magnet. Participants viewed the projection screen through a mirror mounted on the head coil. A practice block of 20 items was completed prior to commencement of the experiment.

Image Acquisition

Images were acquired with a 4-T Bruker MedSpec system using a transverse electromagnetic head coil (Vaughan et al. 2002). A total of 570 echo planar imaging (EPI) images were acquired over 2 runs (time echo [TE] = 30 ms, time repetition [TR] = 2100 ms). Each brain volume consisted of 36 planes, in-plane resolution 3.6 mm, and slice thickness 3 mm (0.6-mm gap). A high-resolution magnetization-prepared rapid gradient-echo (MP-RAGE) 3D T1 image was acquired within the same session (time to inversion [TI] = 700 ms, TR = 1500 ms, TE = 3.35 ms, resolution 0.9 × 0.9 × 0.9 mm). The first 5 volumes in each session were ignored so that magnetization could reach steady state. Foam padding was used to limit head movement.

Image Processing and Analysis

Data were preprocessed and analyzed using statistical parametric mapping (SPM) software (SPM2; Wellcome Department of Cognitive Neurology, Queens Square, London, UK). Data were slice timing corrected, spatially normalized via nonlinear basis function to T1 and EPI template images, and smoothed (full width at half maximum [FWHM] 8 mm Gaussian). A general linear model was applied to the signal intensity time course of each voxel (Worsley and Friston 1995). The fixed-effects model included separate covariates consisting of a synthetic hemodynamic response function (HRF) and its temporal derivative for transient BOLD responses to 5 different trial types: correctly performed trials from dominant, subordinate, unrelated, and nonword conditions and errors. Realignment parameters were used within the design as covariates of no interest. Planned contrasts were employed to compare the HRF parameter estimates only for the experimental conditions for each participant. Drug main effects were examined by collapsing dominant, subordinate, and unrelated conditions. Drug-by-condition interactions were examined for each priming comparison (dominant vs. subordinate, dominant vs. unrelated, and subordinate vs. unrelated). The resulting contrast images from the single within-subject fixed-effects analyses were masked to include only gray matter voxels using the mask provided in SPM2 and entered into 1-tailed t-tests in group random-effects analyses to allow inferences about condition-by-drug interactions and drug main effects across participants.

Following Mehta et al. (2000), for drug-by-condition interactions, voxels exceeding P < 0.001 (uncorrected) with a minimum cluster of 5 voxels are reported for regions where we had reasonable a priori expectations of modulated priming-related activity on L-DOPA (see Introduction). Exploratory whole-brain analyses were also conducted using a significance level of P < 0.05, corrected, as per (Mehta et al. 2000).

Results

Cardiovascular and Mood Effects

Plasma DA levels were increased on L-DOPA (mean = 7145 pmol/L) versus placebo (mean = 302 pmol/L). Paired-sample t-tests showed that there were no significant differences on placebo versus L-DOPA in self-rated subjective mood (Bond and Lader 1974) or in blood pressure or heart rate measured at baseline and immediately after testing (P > 0.05).

Behavioral Effects

Behavioral reaction time (RT) analyses were conducted on correct real-word lexical decisions.

Errors were only made on 1.69% of real-word trials overall, with no main effect of drug (P > 0.13) and no drug-by-condition interaction (P > 0.13). Prior to behavioral latency analysis, outliers (RTs > 2 standard deviation from the mean per the subject condition) were replaced with mean estimator for that condition (1.8% of trials were replaced in this way). The mean correct lexical decision RTs are shown in Table 3 as a function of priming condition and drug. A repeated-measures analysis of variance (ANOVA) was performed with the factors of drug (L-DOPA, placebo) and condition (dominant, subordinate, unrelated). There was no main effect for drug (P = 0.987), but there was a significant main effect of condition (F2,18 = 34.444, P < 0.0001) and a significant drug-by-condition interaction (F2,18 = 3.872, P = 0.040). To further examine this interaction, we conducted repeated-measures ANOVAs on dominant and subordinate priming separately and performed paired-sample t-tests on priming effects (dominant vs. unrelated, subordinate vs. unrelated) separately on L-DOPA and on placebo. A repeated-measures ANOVA on dominant priming (i.e., dominant and unrelated conditions) indicated a main effect of condition (F1,19 = 66.764, P < 0.0001) but no significant main effect for drug (P = 0.618) or drug-by-condition interaction (P = 0.699). Paired-sample t-tests indicated significant dominant priming both on L-DOPA and placebo (P < 0.0001). A repeated-measures ANOVA on subordinate priming effects (i.e., subordinate and unrelated conditions) indicated a main effect for condition (F1,19 = 14.320, P = 0.001) and a marginal drug-by-condition interaction (F1,19 = 4.062, P = 0.058), with paired-sample t-tests indicating significant subordinate priming on placebo (P = 0.001) but no significant subordinate priming on L-DOPA (P > 0.05).

Table 3

Mean lexical decision latencies (in ms) as a function of priming condition and drug

Drug Dominant Subordinate Unrelated 
Placebo 547 (138) 564 (140) 592 (161) 
L-DOPA 539 (121) 576 (143) 588 (138) 
Drug Dominant Subordinate Unrelated 
Placebo 547 (138) 564 (140) 592 (161) 
L-DOPA 539 (121) 576 (143) 588 (138) 

Note: Standard deviations are in parentheses.

Imaging Results

Drug-by-condition interactions were investigated for each priming comparison (see Table 4 and Figs 1 and 2) using a threshold of P < 0.001 uncorrected for regions where we had reasonable a priori expectations of DAergic modulation of priming (see Introduction), as per Mehta et al (2000). Significant interactions were further explored by first calculating for each subject the mean signal in a 6-mm sphere centered on the peak maxima for each condition using Marsbar software (http://www.marsbar.sourceforge.net) and then conducting paired t-tests on the conditions showing a significant interaction separately for L-DOPA and placebo conditions.

Table 4

Regions showing a significant condition-by-drug interaction

Region Z score Coordinates
 
P 
x y z  
Dominant versus subordinate      
    L anterior cingulate 3.41 −3 33 21 0.000 
    R anterior cingulate 3.33 36 21 0.000 
Subordinate versus unrelated      
    L middle frontal gyrus 3.84 −24 24 57 0.000 
    L superior frontal gyrus 3.57 −18 15 63 0.000 
Dominant versus unrelated      
    L rolandic operculum 4.01 −63 −3 0.000 
    L middle temporal gyrus 3.58 −57 −48 0.000 
Region Z score Coordinates
 
P 
x y z  
Dominant versus subordinate      
    L anterior cingulate 3.41 −3 33 21 0.000 
    R anterior cingulate 3.33 36 21 0.000 
Subordinate versus unrelated      
    L middle frontal gyrus 3.84 −24 24 57 0.000 
    L superior frontal gyrus 3.57 −18 15 63 0.000 
Dominant versus unrelated      
    L rolandic operculum 4.01 −63 −3 0.000 
    L middle temporal gyrus 3.58 −57 −48 0.000 

Note: L, left; R, right.

Figure 1.

Regions showing a significant priming-by-drug interaction shown on slices from a canonical T1 high-resolution image included in SPM (P < 0.001, uncorrected with a spatial extent minimum of 5 contiguous voxels per cluster).

Figure 1.

Regions showing a significant priming-by-drug interaction shown on slices from a canonical T1 high-resolution image included in SPM (P < 0.001, uncorrected with a spatial extent minimum of 5 contiguous voxels per cluster).

Figure 2.

Bar graphs indicating mean percent BOLD signal change across subjects (n = 20) as a function of drug and priming condition (error bars indicate standard error). White = dominant, gray = subordinate, black = unrelated.

Figure 2.

Bar graphs indicating mean percent BOLD signal change across subjects (n = 20) as a function of drug and priming condition (error bars indicate standard error). White = dominant, gray = subordinate, black = unrelated.

A significant drug-by-condition interaction was obtained for the dominant versus unrelated contrast in the left rolandic operculum, where there was a significantly increased signal for the dominant condition relative to the unrelated condition on L-DOPA (P = 0.025), whereas on placebo, there was marginal decrease in signal for the dominant condition relative to the unrelated condition (P = 0.06). There was also a significant drug-by-condition interaction for the dominant versus unrelated conditions in the left middle temporal gyrus, where again there was a significant increase in signal for the dominant condition compared with the unrelated condition (P = 0.043) but no difference on placebo. A significant drug-by-condition interaction was observed for the subordinate versus unrelated contrast in the left MFG and adjacent superior frontal gyrus. In the left MFG, the BOLD signal for the subordinate condition was significantly greater than the unrelated condition on placebo (P = 0.029) but significantly reduced on L-DOPA (P = 0.012). There were no significant pairwise differences in the left superior frontal gyrus. A direct comparison of dominant and subordinate conditions showed a significant drug-by-condition interaction in the anterior cingulate bilaterally. There were no significant pairwise differences in BOLD signal on placebo, whereas on L-DOPA, there was a significant difference between dominant and subordinate conditions (reflecting an increase in dominant condition signal and a decrease in the subordinate condition) that was significant in the right anterior cingulate cortex (ACC) (P = 0.009), and this difference was marginal in the left ACC (P = 0.09). No significant drug-by-condition interactions were observed in the exploratory whole-brain analysis with a threshold of P < 0.05 corrected for multiple comparisons using the false discovery rate method (Genovese et al. 2002).

Examination of the main effect of drug across the 3 critical conditions (see Table 5) showed significantly increased BOLD signal on L-DOPA compared with placebo in the left superior temporal gyrus, left hippocampus, and the insula bilaterally (P < 0.001, uncorrected). Decreased activity was observed on L-DOPA versus placebo in the right parahippocampal complex (P < 0.001, uncorrected).

Table 5

Regions showing a main effect for L-DOPA versus placebo

Region Z score Coordinates
 
P 
x y z  
L-DOPA > placebo      
    Left superior temporal gyrus 3.74 −54 −45 15 0.000 
    Right insula 3.66 42 12 −9 0.000 
    Left hippocampus 3.59 −30 −27 −12 0.000 
    Left insula 3.53 −36 15 −12 0.000 
L-DOPA < placebo      
    Right parahippocampal complex 3.71 21 −6 −24 0.000 
Region Z score Coordinates
 
P 
x y z  
L-DOPA > placebo      
    Left superior temporal gyrus 3.74 −54 −45 15 0.000 
    Right insula 3.66 42 12 −9 0.000 
    Left hippocampus 3.59 −30 −27 −12 0.000 
    Left insula 3.53 −36 15 −12 0.000 
L-DOPA < placebo      
    Right parahippocampal complex 3.71 21 −6 −24 0.000 

Regions showing a significant (P < 0.001, uncorrected) main effect for priming across L-DOPA and placebo conditions are presented in Table 6. Dominant priming (dominant vs. unrelated) was associated with increased activity in the inferior frontal gyrus bilaterally and the right supramarginal gyrus and middle cingulate, whereas there was decreased activity for the dominant condition versus the unrelated condition in the right superior temporal gyrus. A main effect for subordinate priming was observed as increased left thalamic activity and reduced left MFG activity for the subordinate condition relative to the unrelated condition. The direct contrast of dominant versus subordinate conditions across drug treatments showed a significant increase for the dominant condition in the left superior occipital lobe, MFG, superior temporal gyrus, parahippocampal gyrus, and middle cingulate in addition to increases in the right thalamus and middle temporal gyrus.

Table 6

Regions showing a significant main effect for priming condition

Region Z score Coordinates
 
P 
x y z  
Dominant > unrelated      
    L inferior frontal gyrus, operculum 4.35 −36 27 0.000 
    R supramarginal gyrus 3.89 60 −27 21 0.000 
    R inferior frontal gyrus, operculum 3.76 57 18 15 0.000 
    R middle cingulate 3.52 −24 39 0.000 
Dominant < unrelated      
    R superior temporal gyrus 3.47 54 −9 −15 0.000 
Subordinate > unrelated      
    L thalamus 3.60 −9 −9 12 0.000 
Subordinate < unrelated      
    L middle frontal gyrus 3.59 −33 15 51 0.000 
Dominant > subordinate      
    L superior occipital 4.56 −18 −69 27 0.000 
    R thalamus 4.06 −18 −3 0.000 
    L middle frontal gyrus 3.79 −27 12 48 0.000 
    L superior temporal gyrus 3.58 −63 −45 18 0.000 
    L parahippocampal gyrus 3.47 −24 −36 −9 0.000 
    R middle temporal gyrus 3.43 60 −27 −9 0.000 
    L middle cingulate 3.31 −9 −27 36 0.000 
Region Z score Coordinates
 
P 
x y z  
Dominant > unrelated      
    L inferior frontal gyrus, operculum 4.35 −36 27 0.000 
    R supramarginal gyrus 3.89 60 −27 21 0.000 
    R inferior frontal gyrus, operculum 3.76 57 18 15 0.000 
    R middle cingulate 3.52 −24 39 0.000 
Dominant < unrelated      
    R superior temporal gyrus 3.47 54 −9 −15 0.000 
Subordinate > unrelated      
    L thalamus 3.60 −9 −9 12 0.000 
Subordinate < unrelated      
    L middle frontal gyrus 3.59 −33 15 51 0.000 
Dominant > subordinate      
    L superior occipital 4.56 −18 −69 27 0.000 
    R thalamus 4.06 −18 −3 0.000 
    L middle frontal gyrus 3.79 −27 12 48 0.000 
    L superior temporal gyrus 3.58 −63 −45 18 0.000 
    L parahippocampal gyrus 3.47 −24 −36 −9 0.000 
    R middle temporal gyrus 3.43 60 −27 −9 0.000 
    L middle cingulate 3.31 −9 −27 36 0.000 

Note: L, left; R, right.

Discussion

To our knowledge, this study provides the first evidence of the brain mechanisms associated with DAergic modulation of semantic processing. The behavioral effect of L-DOPA was to extinguish priming of subordinate meanings, consistent with a DA-induced enhancement of semantic salience or, from an information processing perspective, focusing of activation in semantic networks through inhibition or restricted spreading of activation to weaker representations. We expected priming-related modulation of the network normally engaged during automatic semantic priming on L-DOPA. Priming of dominant meanings was not significantly altered on L-DOPA; however, there was increased activity in the left rolandic operculum and the middle temporal gyrus, with the latter involvement consistent with modulation of automatic semantic priming mechanisms. The lack of behavioral priming for subordinate meanings on L-DOPA was associated with decreased anterior cingulate and dorsal prefrontal cortical activity, suggesting that L-DOPA enhances frequency-based semantic focus via prefrontal modulation of automatic semantic priming and through engagement of anterior cingulate mechanisms supporting attentional/controlled priming.

The clear effect of L-DOPA on behavioral priming was to reduce activation of subordinate or weaker competing representations, reflecting increasing semantic focus or salience on the basis of meaning frequency, as per our earlier behavioral data (Copland, Chenery, et al. 2003). This finding suggests the need to reconceptualize earlier proposals that DA is called upon during lexical disambiguation primarily when internal representations of context are required and dominant meaning biases must be overcome (Cohen and Servan-Schreiber 1992). Instead, it appears that in the absence of context, increasing DAergic transmission may still enhance selection from competing representations but in this instance via the heightening of frequency-based differences. This finding provides further evidence that DA acts to increase the SNR within semantic networks by increasing focus on relevant representations and reducing activity for distracting or irrelevant representations, consistent with the information processing account proposed by Kischka et al. (1996). The current interpretation is similarly based on the assumption that a DA-induced increase in SNR within semantic networks occurs on the basis of meaning frequency when ambiguities are presented without context. The inclusion of a neutral condition in future studies may allow us to further determine the role of DA in selecting and/or inhibiting semantic representations.

The behavioral pattern of results appears consistent with direct evidence from animal studies (Seamans and Yang 2004) and complementary proposals in humans (Cools 2006) that DA serves to sharpen and stabilize representations in the prefrontal cortex by inhibiting weaker representations and narrowing the focus to a single representation or fewer representations that are subsequently maintained. L-DOPA is taken up into DA neurons, converted to DA, and released from vesicles into the synaptic cleft in a stimulation-dependent manner. Its administration may thus modulate DA neurons by 2 independently regulated DA release states (Grace 1991), 1) a tonic release of DA from DA terminals that regulate slow irregular (steady state) firing that may maintain alertness during learning and may influence some aspects of working memory and 2) a rapid phasic release of DA induced by activation of the DA neurons in a stimulus-dependent manner (occurring approximately 80–150 ms after stimulus onset), leading to burst firing that can signal stimulus salience (Schultz 2002). Neuromodulation of alertness, maintenance of representations, and rapid signaling of salience provide plausible mechanisms accounting for the changes in semantic priming we observed on L-DOPA. The precise manner in which these DA transmission mechanisms influence semantic priming is presently unclear.

Given that the present study employed a lexical decision priming task with a short SOA, it was expected that L-DOPA would modulate activity within the network of regions typically involved in automatic semantic priming (see Table 1). Indeed, priming of dominant meanings was associated in increased activity in the left middle temporal gyrus on L-DOPA compared with placebo, suggesting increased engagement of the neural mechanisms underpinning automatic semantic priming (Copland, de Zubicaray, et al. 2003) and consistent with evidence of DA receptors in the temporal lobe (Goldsmith and Joyce 1996). The left MFG has also been implicated in semantic priming with a short SOA on several occasions (Kotz et al. 2002; Rissman et al. 2003). In the present study, behavioral priming of subordinate meanings on placebo was associated with increased left MFG activity for the subordinate versus unrelated conditions. On L-DOPA, there was reduced (i.e., negative) left MFG activity for the subordinate condition compared with a positive signal for the unrelated condition. The reduced activity for the subordinate condition on L-DOPA in this region may reflect a direct modulation of automatic priming in the form of reduced spreading of activation to weaker meanings. However, the change from a positive to negative BOLD signal for the subordinate condition may be more consistent with L-DOPA–based suppression of weaker meanings. The involvement of the left MFG in the DAergic modulation of priming observed here is consistent with strong evidence from human and animal studies that the prefrontal cortex (especially the MFG) is a key substrate for the DAergic influences on cognition (Arnsten 1998; Kimberg et al. 2001; Mattay et al. 2003; Williams and Castner 2006) but further extends this modulatory role to semantics.

The second mechanism by which L-DOPA may influence priming in the present study is to engage neural mechanisms supporting attentional or strategic forms of priming. In the absence of drugs, selective priming of the dominant and not subordinate condition is typically witnessed at longer SOAs, (Simpson and Burgess 1985), where it is assumed that attentional mechanisms are usually engaged over time to inhibit subordinate meanings and sustain dominant meaning activation. The observation of this priming pattern on L-DOPA at a short SOA in the present study supports the notion that increased DA accelerates lexical processing, consistent with previous evidence of earlier semantic activation in healthy individuals on L-DOPA compared with placebo (Angwin et al. 2004) and the finding that lexical activation is slowed in PD where DA deafferentation occurs (Angwin, Copland, et al. 2006). A systematic examination of the time course of lexical activation on L-DOPA with multiple SOAs and noncompeting lexical representations would allow further testing of this proposal and whether such speeded lexical activation occurs when semantic salience is not explicitly manipulated. Whereas a neutral condition is needed to verify the presence of inhibition, the increased semantic selectivity observed here on L-DOPA suggests that DA facilitates increased and earlier attentional processing. Such an effect is consistent with previous evidence that DA enhances attentional capacity during information processing (Clark et al. 1987) and that enhanced DAergic transmission improves performance when attentional mechanisms are required to extract relevant information in situations of representational conflict (Cohen and Servan-Schreiber 1992).

The influence of L-DOPA on attentional mechanisms is also suggested by increased activity for the primed dominant condition and decreased activity for the subordinate condition observed in the anterior cingulate, a region that did not appear to be involved in priming on placebo, as per our predictions. Posner and DiGirolamo (1998) proposed that the anterior cingulate is responsible for “top–down” attentional processes required for selection among competing representations, including the selection of one meaning for lexical ambiguities (but see Botvinick et al. 2001). Whereas the lexical decision priming paradigm employed here does not require active meaning selection for successful task completion, the preserved and selective priming of the dominant condition observed demonstrates that meaning selection had occurred at the semantic level under L-DOPA. In terms of attentional priming mechanisms, expectancy-based priming is unlikely to develop with such a short SOA (Neely 1991); however, controlled postlexical semantic matching processes, where the subject looks for a semantic relationship between the target and the prime, are possible within this time frame (Degroot 1985). Critically, the anterior cingulate has been associated with strategic inhibitory mechanisms in semantic priming (Mummery et al. 1999; Gold et al. 2006) and in strategic lexical ambiguity priming employing the same paradigm as the present study with a longer SOA (Copland et al. 2007). The present findings suggest that L-DOPA facilitates anterior cingulate–based attentional priming mechanisms that are not engaged on placebo, as demonstrated by the lack of anterior cingulate involvement on placebo in the present study.

The ACC is richly innervated by DA projections (Gaspar et al. 1989). Furthermore, DAergic modulation of cognitive function has been linked to the anterior cingulate. For instance, increased anterior cingulate activity has been observed in response to DAergic augmentation in individuals with schizophrenia during verbal fluency tasks (Dolan et al. 1995), and anterior cingulate activity is modulated during working memory performance in individuals with PD on versus off L-DOPA (Mattay et al. 2002). There is also positron emission tomography evidence consistent with increased DA release in the anterior cingulate during working memory and attentional tasks in healthy individuals (Aalto et al. 2005). L-DOPA may engage the anterior cingulate that enables increased attentional processing during the semantic priming task, resulting in accelerated meaning selection and suppression. Comparison of automatic and controlled aspects of semantic priming under L-DOPA may shed further light on the attentional role of the anterior cingulate.

With regard to regions associated with L-DOPA–induced changes in behavioral priming, modulated anterior cingulate and prefrontal activity may arise through 2 distinct mechanisms. The drug may affect brain regions rich in DA receptors directly or have an indirect influence on areas with few DA receptors “downstream” from the former regions (Kimberg et al. 2001). In the present case, it appears more likely that L-DOPA modulated neuronal activity in these regions directly, given evidence of substantial DA receptors in anterior cingulate and prefrontal cortical regions and the fact that in those conditions where behavioral changes were witnessed on L-DOPA changes in brain activity were not observed “upstream” in regions more densely populated with DA receptors. However, a limitation of the current study is that DAergic transmission was not directly measured.

It is also important to consider whether the current findings reflect generalized vasodilatory effects of L-DOPA on cerebral blood flow. We argue that this is unlikely on several grounds. First, the regions showing a significant drug-by-condition interaction relevant to the behavioral effects of L-DOPA did not overlap with regions showing a main effect for drug. Second, the drug-by-priming interactions reflected both increased and decreased BOLD signal, depending on the priming condition. The interactions observed are, therefore, interpreted as possibly reflecting the effects of L-DOPA on local neuronal firing in particular task-relevant brain regions, as per other imaging studies of L-DOPA with similar findings against generalized vascular mechanisms (Cools et al. 2002). Finally, it is acknowledged that L-DOPA is also a precursor to norepinephrine; however, it is argued that the regions identified in the present study as being involved in the modulation of semantic processing are more strongly linked to DAergic than noradrenergic systems (see also Kischka et al. 1996).

Overall, the present findings are consistent with DAergic enhancement of SNRs within semantic networks through modulating prefrontal and temporal mechanisms involved in automatic priming and via engagement of attentional mechanisms that may support the suppression of weaker competing representations. Future imaging studies of priming with selective DA agonists and with other DA imaging techniques will provide further insights into the neural basis of these modulatory effects. The current findings provide some clues regarding the neural mechanisms underpinning altered semantic processing in populations with DAergic dysregulation, including individuals with schizophrenia (Goldberg et al. 2000) and PD (Copland 2003).

Funding

Australian Research Council (discovery grant DP0452264 to D.A.C., K.M.; research fellowship to D.A.C.).

Authors D.C., K.M., P.S., and G.Z. reported no biomedical financial interests. We thank Jenny Stephenson and Martina Bryant for assistance in conducting the study. Conflict of Interest: None declared.

References

Aalto
S
Bruck
A
Laine
M
Nagren
K
Rinne
JO
Frontal and temporal dopamine release during working memory and attention tasks in healthy humans: a positron emission tomography study using the high-affinity dopamine D-2 receptor ligand [C-11]FLB 457
J Neurosci
 , 
2005
, vol. 
25
 (pg. 
2471
-
2477
)
Angwin
AJ
Chenery
HJ
Copland
DA
Arnott
WL
Grattan
R
Murdoch
BE
Silburn
PA
Priming of semantic features in Parkinson's disease
Brain Lang
 , 
2006
, vol. 
99
 (pg. 
102
-
103
)
Angwin
AJ
Chenery
HJ
Copland
DA
Arnott
WL
Murdoch
BE
Silburn
PA
Dopamine and semantic activation: an investigation of masked direct and indirect priming
J Int Neuropsychol Soc
 , 
2004
, vol. 
10
 (pg. 
15
-
25
)
Angwin
AJ
Copland
DA
Chenery
HJ
Murdoch
BE
Silburn
PA
The influence of dopamine on semantic activation in Parkinson's disease: evidence from a multipriming task
Neuropsychology
 , 
2006
, vol. 
20
 (pg. 
299
-
306
)
Arnsten
AFT
Catecholamine modulation of prefrontal cortical cognitive function
Trends Cogn Sci
 , 
1998
, vol. 
2
 (pg. 
436
-
447
)
Balota
DAC
Cortese
M
Pilotti
M
Item-level analyses of lexical decision performance: results from a mega-study
Abstracts of the 40th Annual Meeting of the Psychonomics Society; Los Angeles, CA
 , 
1999
Los Angeles (CA)
Psychonomic Society
pg. 
44
 
Bond
A
Lader
M
Use of analog scales in rating subjective feelings
Br J Med Psychol
 , 
1974
, vol. 
47
 (pg. 
211
-
218
)
Botvinick
MM
Braver
TS
Barch
DM
Carter
CS
Cohen
JD
Conflict monitoring and cognitive control
Psychol Rev.
 , 
2001
, vol. 
108
 (pg. 
624
-
652
)
Clark
CR
Geffen
GM
Geffen
LB
Catecholamines and attention. 2. Pharmacological studies in normal humans
Neurosci Biobehav Rev
 , 
1987
, vol. 
11
 (pg. 
353
-
364
)
Cohen
JD
Servan-Schreiber
D
Context, cortex, and dopamine—a connectionist approach to behavior and biology in schizophrenia
Psychol Rev
 , 
1992
, vol. 
99
 (pg. 
45
-
77
)
Cools
R
Dopaminergic modulation of cognitive function-implications for L-DOPA treatment in Parkinson's disease
Neurosci Biobehav Rev
 , 
2006
, vol. 
30
 (pg. 
1
-
23
)
Cools
R
Lewis
SJG
Clark
L
Barker
RA
Robbins
TW
L-DOPA disrupts activity in the nucleus accumbens during reversal learning in Parkinson's disease
Neuropsychopharmacology
 , 
2007
, vol. 
32
 (pg. 
180
-
189
)
Cools
R
Stefanova
E
Barker
RA
Robbins
TW
Owen
AM
Dopaminergic modulation of high-level cognition in Parkinson's disease: the role of the prefrontal cortex revealed by PET
Brain
 , 
2002
, vol. 
125
 (pg. 
584
-
594
)
Copland
D
The basal ganglia and semantic engagement: potential insights from semantic priming in individuals with subcortical vascular lesions, Parkinson's disease, and cortical lesions
J Int Neuropsychol Soc
 , 
2003
, vol. 
9
 (pg. 
1041
-
1052
)
Copland
DA
Chenery
HJ
Murdoch
BE
Arnott
WL
Silburn
PA
Dopamine enhances semantic salience: semantic priming evidence from healthy individuals
Brain Lang
 , 
2003
, vol. 
87
 (pg. 
103
-
104
)
Copland
DA
de Zubicaray
GI
McMahon
K
Eastburn
M
Neural correlates of semantic priming for ambiguous words: an event-related fMRI study
Brain Res
 , 
2007
, vol. 
1131
 (pg. 
163
-
172
)
Copland
DA
de Zubicaray
GI
McMahon
K
Wilson
SJ
Eastburn
M
Chenery
HJ
Brain activity during automatic semantic priming revealed by event-related functional magnetic resonance imaging
NeuroImage
 , 
2003
, vol. 
20
 (pg. 
302
-
310
)
Dale
AM
Optimal experimental design for event-related fMRI
Hum Brain Mapp
 , 
1999
, vol. 
8
 (pg. 
109
-
114
)
Degroot
AMB
Word-context effects in word naming and lexical decision
Q J Exp Psychol A
 , 
1985
, vol. 
37
 (pg. 
281
-
297
)
Dolan
RJ
Fletcher
P
Frith
CD
Friston
KJ
Frackowiak
RSJ
Grasby
PM
Dopaminergic modulation of impaired cognitive activation in the anterior cingulate cortex in schizophrenia
Nature
 , 
1995
, vol. 
378
 (pg. 
180
-
182
)
Gaspar
P
Berger
B
Febvret
A
Vigny
A
Henry
JP
Catecholamine innervation of the human cerebral-cortex as revealed by comparative immunohistochemistry of tyrosine-hydroxylase and dopamine-beta-hydroxylase
J Comp Neurol
 , 
1989
, vol. 
279
 (pg. 
249
-
271
)
Genovese
CR
Lazar
NA
Nichols
T
Thresholding of statistical maps in functional neuroimaging using the false discovery rate
NeuroImage
 , 
2002
, vol. 
15
 (pg. 
870
-
878
)
Gold
BT
Balota
DA
Jones
SJ
Powell
DK
Smith
CD
Andersen
AH
Dissociation of automatic and strategic lexical-semantics: functional magnetic resonance imaging evidence for differing roles of multiple frontotemporal regions
J Neurosci
 , 
2006
, vol. 
26
 (pg. 
6523
-
6532
)
Goldberg
TE
Dodge
M
Aloia
M
Egan
MF
Weinberger
DR
Effects of neuroleptic medications on speech disorganization in schizophrenia: biasing associative networks towards meaning
Psychol Med
 , 
2000
, vol. 
30
 (pg. 
1123
-
1130
)
Goldsmith
SK
Joyce
JN
Dopamine D2 receptors are organized in bands in normal human temporal cortex
Neuroscience
 , 
1996
, vol. 
74
 (pg. 
435
-
451
)
Grace
AA
Phasic versus tonic dopamine release and the modulation of dopamine system responsivity—a hypothesis for the etiology of schizophrenia
Neuroscience
 , 
1991
, vol. 
41
 (pg. 
1
-
24
)
Kimberg
DY
Aguirre
GK
Lease
J
D'Esposito
M
Cortical effects of bromocriptine, a D-2 dopamine receptor agonist, in human subjects, revealed by fMRI
Hum Brain Mapp
 , 
2001
, vol. 
12
 (pg. 
246
-
257
)
Kischka
U
Kammer
T
Maier
S
Weisbrod
M
Thimm
M
Spitzer
M
Dopaminergic modulation of semantic network activation
Neuropsychologia
 , 
1996
, vol. 
34
 (pg. 
1107
-
1113
)
Kotz
SA
Cappa
SF
von Cramon
DY
Friederici
AD
Modulation of the lexical-semantic network by auditory semantic priming: an event-related functional MRI study
NeuroImage
 , 
2002
, vol. 
17
 (pg. 
1761
-
1772
)
Kucera
H
Francis
WN
Computational analysis of present-day American English
 , 
1967
Providence (RI)
Brown University Press
Matsumoto
A
Iidaka
T
Haneda
K
Okada
T
Sadato
N
Linking semantic priming effect in functional MRI and event-related potentials
NeuroImage
 , 
2005
, vol. 
24
 (pg. 
624
-
634
)
Mattay
VS
Goldberg
TE
Fera
F
Hariri
AR
Tessitore
A
Egan
MF
Kolachana
B
Callicott
JH
Weinberger
DR
Catechol O-methyltransferase val(158)-met genotype and individual variation in the brain response to amphetamine
Proc Natl Acad Sci U S A
 , 
2003
, vol. 
100
 (pg. 
6186
-
6191
)
Mattay
VS
Tessitore
A
Callicott
JH
Bertolino
A
Goldberg
TE
Chase
TN
Hyde
TM
Weinberger
DR
Dopaminergic modulation of cortical function in patients with Parkinson's disease
Ann Neurol
 , 
2002
, vol. 
51
 (pg. 
156
-
164
)
Mehta
MA
Owen
AM
Sahakian
BJ
Mavaddat
N
Pickard
JD
Robbins
TW
Methylphenidate enhances working memory by modulating discrete frontal and parietal lobe regions in the human brain
J Neurosci
 , 
2000
, vol. 
20
 (pg. 
1
-
6
)
Mummery
CJ
Shallice
T
Price
CJ
Dual-process model in semantic priming: a functional imaging perspective
NeuroImage
 , 
1999
, vol. 
9
 (pg. 
516
-
525
)
Neely
JH
Besner
D
Humphreys
GW
Semantic priming effects in visual word recognition: a selective review of current findings and theories
Basic processes in reading: visual word recognition
 , 
1991
New Jersey (NJ)
Lawrece Erlbaum
(pg. 
264
-
336
)
Nelson
DL
McEvoy
CL
Walling
JR
Wheeler
JW
The University of South Florida homograph norms
Behav Res Methods Instrum
 , 
1980
, vol. 
12
 (pg. 
16
-
37
)
O'Hare
AJ
Dien
J
Waterson
LD
Savage
CR
Activation of the posterior cingulate by semantic priming: a co-registered ERP/fMRI study
Brain Res.
 , 
2008
, vol. 
1189
 (pg. 
97
-
114
)
Olanow
CW
Schapira
AHV
Rascol
O
Continuous dopamine-receptor stimulation in early Parkinson's disease
Trends Neurosci
 , 
2000
, vol. 
23
 (pg. 
S117
-
S126
)
Posner
MI
DiGirolamo
GJ
Parasuraman
R
Executive attention: conflict, target detection, and cognitive control
The attentive brain
 , 
1998
Cambridge, (MA)
MIT Press
(pg. 
401
-
423
)
Rissman
J
Eliassen
JC
Blumstein
SE
An event-related fMRI investigation of implicit semantic priming
J Cogn Neurosci
 , 
2003
, vol. 
15
 (pg. 
1160
-
1175
)
Robinson
S
Sandstrom
SM
Denenberg
VH
Palmiter
RD
Distinguishing whether dopamine regulates liking, wanting, and/or learning about rewards
Behav Neurosci
 , 
2005
, vol. 
119
 (pg. 
5
-
15
)
Roesch-Ely
D
Weiland
S
Scheffel
H
Schwaninger
M
Hundemer
HP
Kolter
T
Weisbrod
M
Dopaminergic modulation of semantic priming in healthy volunteers
Biol Psychiatry
 , 
2006
, vol. 
60
 (pg. 
604
-
611
)
Rowe
JB
Hughes
L
Ghosh
BCP
Eckstein
D
Williams-Gray
CH
Fallon
S
Barker
RA
Owen
AM
Parkinson's disease and dopaminergic therapy—differential effects on movement, reward and cognition
Brain
 , 
2008
, vol. 
131
 (pg. 
2094
-
2105
)
Schultz
W
Getting formal with dopamine and reward
Neuron
 , 
2002
, vol. 
36
 (pg. 
241
-
263
)
Seamans
JK
Yang
CR
The principal features and mechanisms of dopamine modulation in the prefrontal cortex
Prog Neurobiol
 , 
2004
, vol. 
74
 (pg. 
321
-
321
)
Simpson
GB
Burgess
C
Activation and selection processes in the recognition of ambiguous words
J Exp Psychol Hum Percept Perform
 , 
1985
, vol. 
11
 (pg. 
28
-
39
)
Tivarus
ME
Ibinson
JW
Hillier
A
Schmalbrock
P
Bevesdorf
DQ
An fMRI study of semantic priming: modulation of brain activity by varying semantic distances
Cogn Behav Neurol
 , 
2006
, vol. 
19
 (pg. 
194
-
201
)
Twilley
LC
Dixon
P
Taylor
D
Clark
K
University of Alberta norms of relative meaning frequency for 566 homographs
Mem Cognit
 , 
1994
, vol. 
22
 (pg. 
111
-
126
)
Vaughan
JT
Adriany
G
Garwood
M
Yacoub
E
Duong
T
DelaBarre
L
Andersen
P
Ugurbil
K
Detunable transverse electromagnetic (TEM) volume coil for high-field NMR
Magn Reson Med
 , 
2002
, vol. 
47
 (pg. 
990
-
1000
)
Williams
GV
Castner
SA
Under the curve: critical issues for elucidating D1 receptor function in working memory
Neuroscience
 , 
2006
, vol. 
139
 (pg. 
263
-
276
)
Worsley
KJ
Friston
KJ
Analysis of fMRI time-series revisited—again
NeuroImage
 , 
1995
, vol. 
2
 (pg. 
173
-
181
)