Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Top-Down Modulations from Dorsal Stream in Lexical Recognition: An Effective Connectivity fMRI Study

  • Yuan Deng ,

    Contributed equally to this work with: Yuan Deng, Ruifang Guo

    Affiliation Key Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy of Science, Beijing, China

  • Ruifang Guo ,

    Contributed equally to this work with: Yuan Deng, Ruifang Guo

    Affiliations State Key Lab of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China, Changzhi Medical College, Changzhi, Shanxi, China

  • Guosheng Ding ,

    dinggsh@bnu.edu.cn (GD); pdl3507@bnu.edu.cn (DP)

    Affiliation State Key Lab of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China

  • Danling Peng

    dinggsh@bnu.edu.cn (GD); pdl3507@bnu.edu.cn (DP)

    Affiliation State Key Lab of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China

Top-Down Modulations from Dorsal Stream in Lexical Recognition: An Effective Connectivity fMRI Study

  • Yuan Deng, 
  • Ruifang Guo, 
  • Guosheng Ding, 
  • Danling Peng
PLOS
x

Abstract

Both the ventral and dorsal visual streams in the human brain are known to be involved in reading. However, the interaction of these two pathways and their responses to different cognitive demands remains unclear. In this study, activation of neural pathways during Chinese character reading was acquired by using a functional magnetic resonance imaging (fMRI) technique. Visual-spatial analysis (mediated by the dorsal pathway) was disassociated from lexical recognition (mediated by the ventral pathway) via a spatial-based lexical decision task and effective connectivity analysis. Connectivity results revealed that, during spatial processing, the left superior parietal lobule (SPL) positively modulated the left fusiform gyrus (FG), while during lexical processing, the left SPL received positive modulatory input from the left inferior frontal gyrus (IFG) and sent negative modulatory output to the left FG. These findings suggest that the dorsal stream is highly involved in lexical recognition and acts as a top-down modulator for lexical processing.

Introduction

According to the well-known visual domain hypothesis [1], when humans process visual stimuli, visual information first arrives at the occipital lobe of the brain, and is then processed via the dorsal and ventral visual pathways. In the ventral pathway, visual information is projected to the temporal lobe for the identification of visual features (“what”). In the dorsal pathway, visual information is projected to the parietal lobe for the spatial/motion analysis (“where”) [2].

Consistent with the what/where visual pathways, the ventral and dorsal visual streams also play important roles during reading. According to a modern vision of the cortical networks for reading [3], the ventral pathway (e.g., occipito-temporal region) is responsible for identifying the word form while the dorsal pathway (e.g., posterior parietal region) plays a role in top-down attention and serial reading. This notion is supported by converging evidence from brain imaging studies. For example, in ventral circuits, the left fusiform gyrus [FG, Broadmans area(BA) 37/19)] performs pattern-based visual analysis specific to a word or word-like stimulus [4][9]. In dorsal circuits, evidence from both normal and abnormal reading suggests that the posterior parietal region is involved in word and non-word reading [10], as well as letter position encoding [11][13]. Impaired function in this region may lead to dyslexia [14][17].

Despite the vast amount of knowledge about the ventral and dorsal streams, little is known about how these two pathways interact, especially in word reading. Evidence from electrophysiological studies in monkeys suggests that the dorsal stream interacts with the frontal region of the brain and modulates the ventral stream to assist in visual object recognition [18][20]. Similar evidence has also been found in humans [21][24], emphasizing the importance of top-down feedback in object recognition. In terms of lexical processing, some researchers suggest that the regions in the dorsal pathway may be involved in sequentially allocating attention along the letters [14][15], [25] and providing top-down feedback to the ventral pathway for letter identification [26]. However, others have found that ventral regions connect to the dorsal area of the brain in a feed-forward way for lexical recognition [10], [27][28]. Further investigation is necessary to clarify this inconsistency.

In the current study, we examined interactions between the dorsal and ventral pathways in lexical recognition and explored how these interactions respond to different cognitive demands. To accomplish this, we adopted a spatial-based lexical decision task and took advantage of the uniqueness of Chinese characters. Generally, Chinese characters consist of one or more radicals (components). The radicals have specific positions in a given character (i.e., left, right, top or bottom of the character) [29][30]. In rare cases, changing the position of radicals of one character (e.g., 呆, /dai1/, meaning “stupid”) creates another meaningful characters (e.g., 杏, /xing4/, meaning “almond”). However, for most Chinese characters, changing the position of the radicals will generate a pseudo-character.

In this study, participants were presented three kinds of stimuli (conditions): 1) true characters (TC); 2) pseudo-characters whose radicals could be rearranged to create a true character (radical-rearranged true characters, RTC); 3) pseudo-characters whose radicals could not be rearranged in any way to create a true character (radical-rearranged pseudo-characters, RPC). The task was to judge whether the radicals of the stimulus could form a true character, regardless of whether this required rearrangement. Because conditions 1 and 2 could belong to the mental lexicon of true characters, while condition 3 could not. So, processing of lexical, which is believed to be mediated by the ventral pathway, was revealed by comparing conditions 1 and 2 to condition 3. On the other hand, conditions 2 and 3 require mental rearrangement of the radicals' positions, while condition 1 does not. Thus, spatial processing, which is believed to be mediated by the dorsal pathway, was revealed by comparing conditions 2 and 3 to condition 1.

Neural activations during the character decision task were acquired by using functional magnetic resonance imaging (fMRI). Effective connectivity analysis was employed, which could reveal intrinsic connections between regions despite different task requirements and modulation effects of external factors [31], [32]. This method allowed us to examine the connections between the dorsal and ventral pathways and to disassociate the modulatory effects of different cognitive processes (lexicality versus spatiality) on these connections.

To our knowledge, this is the first report of such an approach to explore the functions and interactions of dorsal and ventral pathways. If the dorsal stream is involved in top-down modulation, we would predict the existence of an intrinsic connection from the dorsal pathway to the ventral pathway and a direct impact of the ventral stream in lexical recognition. We would also predict that connection patterns between the two streams would be modulated by the different cognitive demands.

Methods

Participants

Participants included 12 adults (5 males and 7 females, aged 18 to 26 years,) who were undergraduate or graduate students at Beijing Normal University. All participants were right-handed and had normal hearing and normal or corrected-to-normal vision. None of the participants had any form or history of reading/learning difficulties. Informed written consent was obtained from the participants before the experiment. This study was approved by the Institutional Review Board of the State Key Laboratory of Cognitive Neuroscience and Learning at Beijing Normal University.

Materials

Three types of stimuli (conditions) were presented, as shown in Figure 1: 1) TC, 2) RTC, and 3) RPC. All stimuli consisted of three radicals. RTC stimuli were created by re-arranging the position of any two radicals in a true character. For example, by switching the position of two upper radicals, the RTC stimulus can be transformed into a true character (背, meaning “back”). For the RPC stimuli, no combinations of the three radicals could create a true character.

thumbnail
Figure 1. Examples of stimuli.

TC, true character; RTC, radical-rearranged true characters; RPC, radical-rearranged pseudo-characters. For RTC, the second character is the true character, created by reversing the upper two radicals. English meaning and Chinese pronunciation (pinyin) for true characters are displayed in parenthesis. “+” and “−” indicate higher or lower involvements of factors.

https://doi.org/10.1371/journal.pone.0033337.g001

The task involved judging whether the radicals of each stimulus could form a true character, regardless of whether they would need to be arranged. Radicals of TC and RTC stimuli can compose a true character (lexical effect) while radicals of RPC cannot. So, responses to conditions 1 and 2 should be “Yes” while response to condition 3 should be “No”. On the other hand, radicals of RTC and RPC stimuli require greater spatial manipulation (spatial effect). Both lexicality and spatiality effects are illustrated in Figure 1, in which “+” and “−” indicate high and low involvement of factors in each of the three conditions.

A total of 135 stimuli were used, including 45 stimuli for each condition. TC and RTC had similar word frequencies (p = 0.945), and the visual complexity of characters, indicated by the number of strokes, was balanced across the three types of stimuli (p = 0.448).

fMRI procedure

During fMRI scanning, an event-related design was adopted, and all trials were presented in a pseudorandom sequence. For each trial, a fixation point was presented for 500 milliseconds, followed by a visual stimulus for 1500 milliseconds. The interstimuli interval (ISI) ranged from 4 to 8 seconds. Participants were asked to determine whether the radicals of each stimulus could compose a true character and to respond using an optical response box. Reaction time was recorded. In addition to the 135 stimuli, 45 null events, in which a blank screen was presented and no response was required, served as a baseline. Three fMRI runs were performed and counterbalanced across participants. The duration of each run was 8 minutes.

Image Acquisition

All images were acquired using a 1.5 Tesla Siemens Trio scanner. Participants lay in the scanner with their head position secured with a specially designed vacuum pillow, holding an optical response box in their hands. The head coil was positioned over the participants' head. Participants viewed visual stimuli that were projected onto a screen via a mirror attached to the inside of the head coil. For functional imaging , a susceptibility weighted single-shot echo planar imaging (EPI) method was used to measure blood oxygenation level-dependent (BOLD) changes. The following scan parameters were used: TE = 40 msec, flip angle = 90°, matrix size = 64×64, field of view = 23 cm, slice thickness = 6 mm with 1.2 mm interval, number of slices = 24, TR = 2000 msec. In addition, a high resolution, T1 weighted 3D image was acquired [Spoiled Gradient Echo (SPGR), TR = 21 ms, TE = 8 msec, flip angle = 19°, matrix size = 169×256, field of view = 25 cm, slice thickness = 1.7 mm, number of slices = 96].

fMRI Data Analysis

Data analysis was performed using Statistical Parametric Mapping 2 (SPM2). Functional images were corrected for differences in slice-acquisition time to the middle volume and were realigned to the first volume in the scanning session using affine transformations. No participant had more than 3.0 mm of movement in any plane. Co-registered images were normalized to the MNI average template (12 linear affine parameters for brain size and position, 8 non-linear iterations and 2×2×2 nonlinear basis functions). Statistical analyses were calculated on the smoothed data (6 mm isotropic Gaussian kernel) with a high pass filter (128 seconds cutoff period) in order to remove signal drift, cardiac and respiratory effects, and other low frequency artifacts.

Data from each participant were entered into a general linear model using an event-related analysis procedure. Each individual event was modeled using a canonical hemodynamic response function (HRF). Parameter estimates from contrasts of the canonical HRF in single participant models were entered into random-effects analysis across all participants to determine whether activation during a contrast was significant. First, all conditions were combined and compared to baseline to reveal the overall activation pattern for the cognitive task. Threshold was set at p<0.05, FDR-corrected, with a cluster size of 10 or greater. Then, differences between each condition were examined using paired t-tests with a statistical threshold of p<0.001, uncorrected, with a cluster size of 10 or greater.

To conduct the connectivity analysis, regions-of-interest (ROIs) were selected as those that showed the highest overlap across the three conditions relative to baseline at group level (p<0.05, FDR-corrected) [32]. Three key regions were identified: the left fusiform gyrus (FG, peak: −38, −54, −22), superior parietal lobule (SPL, peak: −24, −66, 56) and the dorsal aspect of inferior frontal gyrus (IFG, peak: −46, 4, 26). The SPM Marsbar toolbox [33] was used to construct ROI regions and to calculate the mean estimate of the percent signal change for each condition. Next, the data for these regions were submitted to a 3 (regions)×3 (conditions) ANOVA analysis.

Finally, the dynamic causal modeling (DCM) tool in SPM2 [31] was used to estimate the effective connectivity between the three ROIs. Based on the group results from random-effect analyses, a DCM model of the three regions was constructed. In addition to these left hemisphere ROIs, several other regions showed overlapping activations across the three conditions, including bilateral middle occipital gyrus, bilateral medial frontal gyrus, bilateral superior frontal gyrus, left precentral gyrus, left postcentral gyrus, left precunes, and right middle frontal gyrus. These regions were not included in the model, as they have not been implicated in previous connectivity analysis of lexical processing [28], [32], [34], and it was therefore impossible to formulate a priori hypotheses of their connectivity.

Regional responses for each subject in each condition were extracted by calculating the principal eigenvariate across all voxels within a 6-mm sphere, centered on the most significantly activated voxel. Subject-specific maxima were defined operationally as the most significantly activated voxels within 10 mm of the group maximum. Data from one subject were excluded because no significant activation cluster was observed within 12 mm from the group-reference voxel in either IFG or SPL. For each DCM model, full and reciprocal connections were specified between FG and SPL and between SPL and IFG. The connection between FG and IFG was not included because previous evidence has demonstrated that FG and IFG are not directly connected during lexical processing, but instead their interactions are mediated via the temporal region [32], [34]. Input conditions were modeled as exerting direct effects on the FG. Two modulation effects were included: 1) lexicality, including TC and RTC conditions, and 2) spatiality, including RTC and RPC conditions. During modeling, three types of parameters were calculated for each subject: 1) intrinsic connections between regions in the absence of modulating experimental effects; 2) both lexical and spatial modulatory effects on intrinsic connections (i.e., the changes in the intrinsic connectivity between regions induced by the experimental design); and 3) input to FG. After modeling, a one-sample t-test was performed for each interregional coupling to determine whether the across-subject mean differed from zero.

Results

Behavioral results

The average RTs were: TC, 941±215 msec; RTC, 1412±168 msec; RPC, 1494±253 msec. Error rates were 20%, 41%, and 38%, respectively. One-way ANOVAs reveal significant main effects of stimulus type on RT [F (2,10) = 17.196, p<0.005] and error rate [F (2,10) = 29.439, p<0.001]. Post-hoc analyses revealed that participants showed significantly faster RT and lower error rate in the TC condition compared to the RTC and RPC conditions (p's<0.01), which has similar RTs and error rates (p's>0.05). These results suggest that the TC condition was less difficult than the RTC and RPC conditions.

Imaging results

As shown in Table 1, overall, the cognitive task activated a broad neural network involving many regions related to lexical processing, including the bilateral fusiform gyrus, middle occipital gyrus, inferior/middle/medial frontal gyrus, and left inferior/superior parietal lobule. As shown in Table S1 and Figure 2, compared to TC, both RTC and RPC produced greater activation in the bilateral fusiform gyrus, middle occipital gyrus, inferior/middle frontal gyrus, inferior/superior parietal lobule, thalamus, and putamen. Compared to RTC, TC produced greater activation in the left postcentral gyrus. Unlike RPC, TC produced activations in the left fusiform gyrus and right middle occipital gyrus

thumbnail
Figure 2. Brain activation maps for differences between conditions.

TC, true character; RTC, radical-rearranged true characters; RPC, radical-rearranged pseudo-characters. Green: RTC/RPC greater than TC. Red: TC greater than RTC/RPC.

https://doi.org/10.1371/journal.pone.0033337.g002

thumbnail
Table 1. Overlapping activations for all character conditions compared to baseline condition.

https://doi.org/10.1371/journal.pone.0033337.t001

Activation intensity for each ROI in each condition are presented in Figure 3. ANOVA revealed significant main effects of condition and region [Fcondition(2,10) = 8.58, p<0.001; Fregion(2,10) = 5.64 p<0.05], as well as their interaction [Fcondition×region(4,8) = 4.29, p<0.05]. Post-hoc analysis revealed higher activation in FG than SPL or IFG in the TC condition, higher activation in FG and SPL than IFG in the RTC condition, and higher activation in SPL than IFG in the RPC condition.

thumbnail
Figure 3. Percentage signal change in ROIs for each condition.

TC, true character; RTC, radical-rearranged true characters; RPC, radical-rearranged pseudo-characters. FG, fusiform gyrus; IFG, inferior frontal gyrus; SPL, superior parietal lobule.

https://doi.org/10.1371/journal.pone.0033337.g003

In addition, consistent with the behavioral results, both the RTC and RPC conditions induced greater activation than the TC condition in all ROIs. To evaluate the influence of task difficulty, activation differences between TC and RTC/RPC were correlated with corresponding differences in behavioral performance for each ROI. Correlations were not significant in IFG or FG, but were significant in SPL for the TC versus RTC comparisons [r(TC vs. RTC) = 0.58, p<.05]. Therefore, greater activations in SPL may reflect higher spatial processing requirements in the RTC and RPC conditions, but task difficulty may not influence activation of IFG or FG.

As shown in Figure 4A, although both the ventral and dorsal aspect of inferior frontal gyrus showed significant activation at the group level, activations were consistent across subjects only in the dorsal region, which was the reason for selecting this particular area as the ROI for connectivity analysis. As shown in Table 2 and Figure 4B, the DCM results revealed a significant intrinsic connection from SPL to FG and a tendency for a connection from SPL to IFG. Spatial processing demand significantly positively modulated the connection from SPL to FG, while lexical judgment demand significantly negatively modulated the connection of the SPL to FG and showed a trend to positively modulate in the connection from IFG to SPL. The negative connectivity values for the connection from SPL to FG indicate an inverse modulatory effect (i.e., increases in activity within the SPL reduced activity within the FG and vice versa).

thumbnail
Figure 4. Differential modulatory effects between task demands.

A, ROI regions that showed the highest overlap across three conditions comparing to baseline at group level (FDR-corrected, p<0.05); B, Yellow solid line arrows indicate influences modulated by lexical judgment, and blue dashed line arrows indicate influences modulated by spatial processing. FG, fusiform gyrus; IFG, inferior frontal gyrus; SPL, superior parietal lobule.

https://doi.org/10.1371/journal.pone.0033337.g004

thumbnail
Table 2. Intrinsic connections between ROIs and modulatory effects on each connection.

https://doi.org/10.1371/journal.pone.0033337.t002

Discussion

These findings demonstrate a role of the left superior parietal lobule in lexical processing, which has important implications for interactions between the dorsal and ventral pathways in reading. Specifically, the left FG showed the greatest activation in lexical recognition (TC condition), while the left SPL and FG showed greater activation with both lexical recognition and spatial processing (RTC and RPC condition). Effective connectivity analysis further demonstrated an intrinsic connection from SPL (dorsal stream) to FG (ventral stream). Finally, spatial processing demand positively modulated the connection from SPL to FG, while lexical processing demand had a negative modulatory effect on this same connection.

A great deal of evidence suggested that regions in the dorsal pathway (e.g., the posterior parietal lobe) are highly involved in lexical recognition. Effective connectivity studies also revealed that dorsal parietal region is a critical part of the pathway for processing both real and pseudo-words in English [10], and dysfunction of the posterior parietal region may partially underlie the reading deficit of dyslexia in English [14], [25]. Neuro-imaging studies in Chinese reading have also suggested that the left superior parietal region plays a critical role in both Chinese character recognition [35][37] and learning [38]. The current study directly demonstrated that the dorsal pathway significantly interacts with the ventral pathway and is highly involved in Chinese character recognition.

Previous studies have reported conflicting findings about the specific role of the dorsal pathway and its interaction with the ventral pathway in reading. Some studies suggested that regions in the ventral pathway connect to the regions in the dorsal pathway in a feed-forward way during lexical recognition [10], [27][28]. For example, researchers found a significant connection from the left inferior occipital gyrus (LIOG) to the left superior parietal lobule (LSPL) (i.e., “where” pathway) in a chunk decomposition task [28]. In addition, the LIOG→LSPL connection significantly increased when processing real characters and pseudo-characters with tight word form (high perceptual tightness), suggesting involvement of the dorsal parietal region in visual-spatial analysis. However, other researchers suggested that the dorsal pathway plays a role in top-down feedback to the ventral pathway for serial letter-order encoding [12], [14], [25][26] and for generation and perception of visual form [39]. Through effective connectivity analyses, our study suggests that there is an intrinsic connection from SPL (dorsal stream) to FG (ventral stream) in the absence of modulating experimental effects, supporting its crucial role in top-down influence to the ventral pathway.

Moreover, the specific function of the connection from SPL (dorsal) to FG (ventral) depends on task demands (i.e., modulatory effects). For spatial-based processing, SPL presented a strong positive top-down modulation of the occipito-temporal region (i.e., an increase in SPL activity leads to an increase in FG activity). When spatial processing was highly involved (RTC and RPC conditions), the left superior parietal region showed greater activation, supporting the dominant role of the dorsal pathway in spatial processing [2].

During lexical recognition, top-down influence exhibited an inhibitory effect (i.e., an increase in SPL activity caused a decrease rate of change in FG activity). Lexical processing also modulated the connection from the dorsal IFG to the SPL. The involvement of the prefrontal region may elucidate the specific role of top-down modulation from SPL to FG. According to previous studies, the dorsal IFG is mainly involved in phonological manipulation [40][42] and word retrieval/selection [43][45]. Some evidence has also suggested that the fronto-parietal network is responsible for word selection [43], [46][49]. In the current study, phonological processing was not likely involved in this task due to the logographic characteristic of Chinese characters and the task requirements (lexical decision), thus the involvement of dorsal IFG is more likely related to controlled semantic retrieval and selection. Word selection likely involves both excitation of intended words and inhibition of unintended words [43]. Therefore, the negative modulatory effect of SPL on FG may be associated with inhibition of the unintended words.

One explanation for the negative modulation effect happening mainly with lexical recognition rather than with spatial processing may have been the task-specific requirements in the current study. Because the character decision task emphasized the lexical processing, therefore, lexical recognition and lexicon selection played a dominant role across conditions regardless of spatial processing demand.

A recently proposed “interactive account” suggested that the ventral occipito-temporal region acts as an interface linking visual word form processing with processing of other aspects in both bottom-up and top-down directions [26], [50]. In the current study, FG showed the greatest activation to real characters and received top-down modulations from SPL during both spatial and lexical task requirements. Therefore, consistent with the interactive account, the current study provides evidence of effective connectivity, indicating top-down modulation from the posterior parietal region to the ventral occipito-temporal region during visual lexical recognition, and suggesting that the ventral pathway plays a dominant role in lexical recognition.

Behavioral performance showed that conditions requiring radical rearrangement (RTC and RPC) were more difficult, which could also affect SPL activation. So these influences seem to be inevitable and predictable due to the current experimental design. Further studies should attempt to control this potential confound. In addition, as previously mentioned, tasks in current study less relied on phonological processing, so future studies could explore the effect of phonological processing demand on the current network.

Conclusion

By taking advantage of the unique qualities of Chinese characters, the current study investigated interactions between the dorsal and ventral pathways in lexical recognition and explored how these connections could be modulated by lexical and spatial demands. Effective connectivity results revealed that there is an intrinsic connection from the left superior parietal lobule (dorsal pathway) to the left fusiform gyrus (ventral pathway), which is modulated by both spatial and lexical demands. Regions in the dorsal stream receive modulation from the prefrontal region, and in turn, modulate the ventral stream to assist in visual recognition of words. Taken together, the current study suggests that the dorsal stream is highly involved in lexical recognition and acts as a top-down modulator for lexical selection.

Supporting Information

Table S1.

Brain regions showing significant activations for each comparison.

https://doi.org/10.1371/journal.pone.0033337.s001

(DOC)

Author Contributions

Conceived and designed the experiments: DLP GSD RFG. Performed the experiments: RFG. Analyzed the data: YD. Wrote the paper: YD GSD DLP.

References

  1. 1. Goodale MA, Milner AD (1992) Separate visual pathways for perception and action. Trends Neurosci 15: 20–25.MA GoodaleAD Milner1992Separate visual pathways for perception and action.Trends Neurosci152025
  2. 2. Bear MF, Connors B, Paradiso M (2007) Neuroscience: Exploring the brain. Hagerstown: Lippincott Williams & Wilkins. MF BearB. ConnorsM. Paradiso2007Neuroscience: Exploring the brainHagerstownLippincott Williams & Wilkins
  3. 3. Dehaene S (2009) Reading in the brain: The science and evolution of a human invention. London: Viking Penguin. 63 p.S. Dehaene2009Reading in the brain: The science and evolution of a human inventionLondonViking Penguin63
  4. 4. Xue G, Chen CS, Jin Z, Dong Q (2006) Language experience shapes fusiform activation when processing a logographic artificial language: An fMRI training study. NeuroImage 31: 1315–1326.G. XueCS ChenZ. JinQ. Dong2006Language experience shapes fusiform activation when processing a logographic artificial language: An fMRI training study.NeuroImage3113151326
  5. 5. Cohen L, Lehericy S, Chochon F, Lemer C, Rivaud S, et al. (2002) Language-specific tuning of visual cortex? Functional properties of the visual word form area. Brain 125: 1054–1069.L. CohenS. LehericyF. ChochonC. LemerS. Rivaud2002Language-specific tuning of visual cortex? Functional properties of the visual word form area.Brain12510541069
  6. 6. Tarkiainen A, Helenius P, Hansen PC, Cornelissen PL, Salmelin R (1999) Dynamics of letter string perception in the human occipito-temporal cortex. Brain 122: 2119–2132.A. TarkiainenP. HeleniusPC HansenPL CornelissenR. Salmelin1999Dynamics of letter string perception in the human occipito-temporal cortex.Brain12221192132
  7. 7. Puce A, Allison T, Asgari M, Gore JC, McCarthy G (1996) Differential sensitivity of human visual cortex to faces, letter strings, and textures: A functional magnetic resonance imaging study. J Neurosci 16: 5205–5215.A. PuceT. AllisonM. AsgariJC GoreG. McCarthy1996Differential sensitivity of human visual cortex to faces, letter strings, and textures: A functional magnetic resonance imaging study.J Neurosci1652055215
  8. 8. Nobre AC, Allison T, McCarthy G (1994) Word recognition in the human inferior temporal lobe. Nature 372: 260–263.AC NobreT. AllisonG. McCarthy1994Word recognition in the human inferior temporal lobe.Nature372260263
  9. 9. Petersen SE, Fox PT, Snyder AZ, Raichle ME (1990) Activation of extrastriate and frontal cortical areas by visual words and word-like stimuli. Science 249: 1041–1044.SE PetersenPT FoxAZ SnyderME Raichle1990Activation of extrastriate and frontal cortical areas by visual words and word-like stimuli.Science24910411044
  10. 10. Levy J, Pernet C, Treserras S, Boulanouar K, Aubry F, et al. (2009) Testing for the dual-route cascade reading model in the brain: An fMRI effective connectivity account of an efficient reading style. PLoS ONE 4(8): e6675.J. LevyC. PernetS. TreserrasK. BoulanouarF. Aubry2009Testing for the dual-route cascade reading model in the brain: An fMRI effective connectivity account of an efficient reading style.PLoS ONE48e6675
  11. 11. Pammer K, Connell E, Kevan A (2009) Reading and spelling: Using visual sensitivity to explore separate or dual orthographic mechanisms. Perception 39: 387–406.K. PammerE. ConnellA. Kevan2009Reading and spelling: Using visual sensitivity to explore separate or dual orthographic mechanisms.Perception39387406
  12. 12. Pammer K, Hansen P, Holliday I, Cornelissen P (2006) Attentional shifting and the role of the dorsal pathway in visual word recognition. Neuropsychologia 44: 2926–2936.K. PammerP. HansenI. HollidayP. Cornelissen2006Attentional shifting and the role of the dorsal pathway in visual word recognition.Neuropsychologia4429262936
  13. 13. Cornelissen PL, Hansen PC, Gilchrist Id, Cormack F, Essex J, et al. (1998) Coherent motion detection and letter position encoding. Vision Res 38: 2181–2191.PL CornelissenPC HansenId GilchristF. CormackJ. Essex1998Coherent motion detection and letter position encoding.Vision Res3821812191
  14. 14. Vidyasagar TR, Pammer K (2010) Dyslexia: a deficit in visuo-spatial attention, not in phonological processing. Trends Cogn Sci 14: 57–63.TR VidyasagarK. Pammer2010Dyslexia: a deficit in visuo-spatial attention, not in phonological processing.Trends Cogn Sci145763
  15. 15. Vidyasagar TR, Pammer K (1999) Impaired visual search in dyslexia relates to the role of the magnocellular pathway in attention. NeuroReport 10: 1283–1287.TR VidyasagarK. Pammer1999Impaired visual search in dyslexia relates to the role of the magnocellular pathway in attention.NeuroReport1012831287
  16. 16. Cornelissen PL, Hansen PC, Hutton JL, Evangelinou V, Stein JF (1998) Magnocellular visual function and children's single word reading. Vision Res 38: 471–482.PL CornelissenPC HansenJL HuttonV. EvangelinouJF Stein1998Magnocellular visual function and children's single word reading.Vision Res38471482
  17. 17. Cornelissen PL, Hansen PC (1998) Motion detection, letter position encoding and single word reading. Ann Dyslexia 48: 155–188.PL CornelissenPC Hansen1998Motion detection, letter position encoding and single word reading.Ann Dyslexia48155188
  18. 18. Saalmann YB, Pigarev IN, Vidyasagar TR (2007) Neural mechanisms of visual attention: How top-down feedback highlights relevant locations. Science 316: 1612–1615.YB SaalmannIN PigarevTR Vidyasagar2007Neural mechanisms of visual attention: How top-down feedback highlights relevant locations.Science31616121615
  19. 19. Buschman TJ, Miller EK (2007) Top-down and bottom-up control of attention in the prefrontal and posterior parietal cortices. Science 315: 1860–1862.TJ BuschmanEK Miller2007Top-down and bottom-up control of attention in the prefrontal and posterior parietal cortices.Science31518601862
  20. 20. Bullier J (2001) Integrated model of visual processing. Brain Res Rev 36: 96–107.J. Bullier2001Integrated model of visual processing.Brain Res Rev3696107
  21. 21. Farivar R (2009) Dorsal-ventral integration in object recognition. Brain Res Rev 61: 144–153.R. Farivar2009Dorsal-ventral integration in object recognition.Brain Res Rev61144153
  22. 22. Laycock R, Crewther SG, Crewther DP (2007) A role for the ‘magnocellular advantage’ in visual impairments in neurodevelopmental and psychiatric disorders. Neurosci Biobehav Rev 31: 363–376.R. LaycockSG CrewtherDP Crewther2007A role for the ‘magnocellular advantage’ in visual impairments in neurodevelopmental and psychiatric disorders.Neurosci Biobehav Rev31363376
  23. 23. Vidyasagar TR (1999) A neuronal model of attentional spotlight: parietal guiding the temporal. Brain Res Brain Res Rev 30: 66–76.TR Vidyasagar1999A neuronal model of attentional spotlight: parietal guiding the temporal.Brain Res Brain Res Rev306676
  24. 24. Gregoriou GG, Gotts SJ, Zhou HH, Desimone R (2009) High-frequency, long-range coupling between prefrontal and visual cortex during attention. Science 324: 1207–1210.GG GregoriouSJ GottsHH ZhouR. Desimone2009High-frequency, long-range coupling between prefrontal and visual cortex during attention.Science32412071210
  25. 25. Vidyasagar TR, Pammer K (2010) Letter-order encoding is both bottom-up and top-down: Response to Whitney. Trends Cogn Sci 14: 238–239.TR VidyasagarK. Pammer2010Letter-order encoding is both bottom-up and top-down: Response to Whitney.Trends Cogn Sci14238239
  26. 26. Twomey T, Duncan KJK, Price CJ, Devlin JT (2011) Top-down modulation of ventral occipito-temporal responses during visual word recognition. NeuroImage 55: 1242–1251.T. TwomeyKJK DuncanCJ PriceJT Devlin2011Top-down modulation of ventral occipito-temporal responses during visual word recognition.NeuroImage5512421251
  27. 27. Whitney C (2010) Serial letter-order encoding is bottom-up, not top-down: comment on Vidyasagar and Pammer. Trends Cogn Sci 14: 237–238.C. Whitney2010Serial letter-order encoding is bottom-up, not top-down: comment on Vidyasagar and Pammer.Trends Cogn Sci14237238
  28. 28. Wu QY, Wu LL, Luo J (2010) Effective connectivity of dorsal and ventral visual pathways in chunk decomposition. Sci China Life Sci 53: 1474–1482.QY WuLL WuJ. Luo2010Effective connectivity of dorsal and ventral visual pathways in chunk decomposition.Sci China Life Sci5314741482
  29. 29. Ding GS, Peng DL, Taft M (2004) The nature of the mental representation of radicals in Chinese: A priming study. J Exp Psychol-Learn Mem Cogn 30: 530–539.GS DingDL PengM. Taft2004The nature of the mental representation of radicals in Chinese: A priming study.J Exp Psychol-Learn Mem Cogn30530539
  30. 30. Taft M, Zhu X, Peng D (1999) Positional specificity of radicals in Chinese character recognition. J Mem Lang 40: 498–519.M. TaftX. ZhuD. Peng1999Positional specificity of radicals in Chinese character recognition.J Mem Lang40498519
  31. 31. Friston KJ, Harrison L, Penny WD (2003) Dynamic causal modelling. NeuroImage 19: 1273–1302.KJ FristonL. HarrisonWD Penny2003Dynamic causal modelling.NeuroImage1912731302
  32. 32. Bitan T, Booth JR, Choy J, Burman DD, Gitelman DR, et al. (2005) Shifts of effective connectivity within a language network during rhyming and spelling. J Neurosci 25: 5397–5403.T. BitanJR BoothJ. ChoyDD BurmanDR Gitelman2005Shifts of effective connectivity within a language network during rhyming and spelling.J Neurosci2553975403
  33. 33. Brett M, Anton JL, Valabregue R, Poline JB (2002) Region of interest analysis using an SPM toolbox. M. BrettJL AntonR. ValabregueJB Poline2002Region of interest analysis using an SPM toolbox.Presented at the 8th International Conference on Functional Mapping of the Human Brain, June 2–6, 2002, Sendai, Japan. Available on CD-ROM in NeuroImage, 16 (2). Presented at the 8th International Conference on Functional Mapping of the Human Brain, June 2–6, 2002, Sendai, Japan. Available on CD-ROM in NeuroImage, 16 (2).
  34. 34. Fan LY, Lee SH, Chou TL (2010) Interaction between brain regions during semantic processing in Chinese adults. Lang Linguist 11: 159–181.LY FanSH LeeTL Chou2010Interaction between brain regions during semantic processing in Chinese adults.Lang Linguist11159181
  35. 35. Chan ST, Tang SW, Tang KW, Lee WK, Lo SS, et al. (2009) Hierarchical coding of characters in the ventral and dorsal visual streams of Chinese language processing. NeuroImage 48: 423–435.ST ChanSW TangKW TangWK LeeSS Lo2009Hierarchical coding of characters in the ventral and dorsal visual streams of Chinese language processing.NeuroImage48423435
  36. 36. Kuo WJ, Yeh TC, Lee CY, Wu YT, Chou CC, et al. (2003) Frequency effects of Chinese character processing in the brain: An event-related fMRI study. NeuroImage 18: 720–730.WJ KuoTC YehCY LeeYT WuCC Chou2003Frequency effects of Chinese character processing in the brain: An event-related fMRI study.NeuroImage18720730
  37. 37. Kuo WJ, Yeh TC, Lee JR, Chen LF, Lee PL, et al. (2004) Orthographic and phonological processing of Chinese characters: An fMRI study. NeuroImage 21: 1721–1731.WJ KuoTC YehJR LeeLF ChenPL Lee2004Orthographic and phonological processing of Chinese characters: An fMRI study.NeuroImage2117211731
  38. 38. Deng Y, Booth J, Chou TL, Ding GS, Peng DL (2008) Item-specific and generalization effects on brain activation when learning Chinese characters. Neuropsychologia 46: 1864–1876.Y. DengJ. BoothTL ChouGS DingDL Peng2008Item-specific and generalization effects on brain activation when learning Chinese characters.Neuropsychologia4618641876
  39. 39. Cardin V, Friston KJ, Zeki S (2011) Top-down modulations in the visual form pathway revealed with dynamic causal modeling. Cereb Cortex 21: 550–562.V. CardinKJ FristonS. Zeki2011Top-down modulations in the visual form pathway revealed with dynamic causal modeling.Cereb Cortex21550562
  40. 40. Devlin JT, Matthews PM, Rushworth MFS (2003) Semantic processing in the left inferior prefrontal cortex: A combined functional magnetic resonance imaging and transcranial magnetic stimulation study. J Cogn Neurosci 15: 71–84.JT DevlinPM MatthewsMFS Rushworth2003Semantic processing in the left inferior prefrontal cortex: A combined functional magnetic resonance imaging and transcranial magnetic stimulation study.J Cogn Neurosci157184
  41. 41. Poldrack RA, Wagner AD, Prull MW, Desmond JE, Glover GH, et al. (1999) Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex. NeuroImage 10: 15–35.RA PoldrackAD WagnerMW PrullJE DesmondGH Glover1999Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex.NeuroImage101535
  42. 42. Fiez JA, Balota DA, Raichle ME, Petersen SE (1999) The effects of word frequency and spelling-to-sound consistency on the functional anatomy of reading. Neuron 24: 205–218.JA FiezDA BalotaME RaichleSE Petersen1999The effects of word frequency and spelling-to-sound consistency on the functional anatomy of reading.Neuron24205218
  43. 43. Price CJ (2010) The anatomy of language: A review of 100 fMRI studies. Ann NY Acad Sci 1191: 62–88.CJ Price2010The anatomy of language: A review of 100 fMRI studies.Ann NY Acad Sci11916288
  44. 44. Heim S, Eickhoff SB, Friederici AD, Amunts K (2009) Left cytoarchitectonic area 44 supports selection in the mental lexicon during language production. Brain Struct Funct 213: 441–456.S. HeimSB EickhoffAD FriedericiK. Amunts2009Left cytoarchitectonic area 44 supports selection in the mental lexicon during language production.Brain Struct Funct213441456
  45. 45. Thompson-Schill SL, D'Esposito M, Aguirre GK, Farah MJ (1997) Role of left inferior prefrontal cortex in retrieval of semantic knowledge: A reevaluation. Proc Natl Acad Sci USA 94: 14792–14797.SL Thompson-SchillM. D'EspositoGK AguirreMJ Farah1997Role of left inferior prefrontal cortex in retrieval of semantic knowledge: A reevaluation.Proc Natl Acad Sci USA941479214797
  46. 46. Tremblay P, Gracco VL (2010) On the selection of words and oral motor responses: Evidence of a response-independent fronto-parietal network. Cortex 46: 15–28.P. TremblayVL Gracco2010On the selection of words and oral motor responses: Evidence of a response-independent fronto-parietal network.Cortex461528
  47. 47. Grindrod CM, Bilenko NY, Myers EB, Blumstein SE (2008) The role of the left inferior frontal gyrus in implicit semantic competition and selection: An event-related fMRI study. Brain Res 1229: 167–178.CM GrindrodNY BilenkoEB MyersSE Blumstein2008The role of the left inferior frontal gyrus in implicit semantic competition and selection: An event-related fMRI study.Brain Res1229167178
  48. 48. Bunge SA, Hazeltine E, Scanlon MD, Rosen AC, Gabrieli JD (2002) Dissociable contributions of prefrontal and parietal cortices to response selection. NeuroImage 17: 1562–1571.SA BungeE. HazeltineMD ScanlonAC RosenJD Gabrieli2002Dissociable contributions of prefrontal and parietal cortices to response selection.NeuroImage1715621571
  49. 49. Gold BT, Buckner RL (2002) Common prefrontal regions coactivate with dissociable posterior regions during controlled semantic and phonological tasks. Neuron 35: 803–812.BT GoldRL Buckner2002Common prefrontal regions coactivate with dissociable posterior regions during controlled semantic and phonological tasks.Neuron35803812
  50. 50. Price CJ, Devlin JT (2011) The interactive account of ventral occipitotemporal contributions to reading. Trends Cog Sci 15: 246–253.CJ PriceJT Devlin2011The interactive account of ventral occipitotemporal contributions to reading.Trends Cog Sci15246253