Transcranial Ultrasound Stimulation Unravels the Causal Role of the Left Inferior Frontal Gyrus in Syntactic Processing

Transcranial Ultrasound Stimulation Unravels the Causal Role of the Left Inferior Frontal Gyrus in Syntactic Processing | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Transcranial Ultrasound Stimulation Unravels the Causal Role of the Left Inferior Frontal Gyrus in Syntactic Processing harry-luyao Chen, Guang Yang, Xingfang Qu, Dongwei Li, Chenguang Zhao, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8717654/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Hierarchical syntactic processing is a fundamental ability for human language processing. Previous studies dispute the left inferior frontal gyrus (LIFG) and the left posterior temporal lobe (LpTL as its possible neural underpinnings, however, direct comparisons of the role of these regions during sentence comprehension are rare. We are among the first to use transcranial ultrasound stimulation (TUS), a non-invasive brain stimulation technique with high precision, to test the causal roles of the different brain regions in processing complex sentences (with embedded relative clauses) and simpler coordinated sentences. Participants completed three lab visits (7 days apart), with LIFG, LpTL, or vertex (control) stimulated each time. Our TUS experiment provides consistent evidence for a causal role of the LIFG in syntactic processing: LIFG TUS significantly decreased d-prime (d’) sensitivity, prolonged reaction times (RTs), and impaired evidence accumulation during decision-making revealed by the hierarchical drift-diffusion modeling. Our findings besides demonstrating the TUS efficacy in sentence-level syntactic processing strengthen the crucial role of LIFG (not LpTL) for hierarchical syntactic processing, thereby deepen the understanding of this core language faculty’s causal neural mechanisms. Biological sciences/Neuroscience/Cognitive neuroscience/Language Biological sciences/Psychology/Human behaviour Sentence processing Posterior temporal lobe Hierarchical syntactic processing Inferior frontal gyrus Transcranial ultrasound stimulation Figures Figure 1 Figure 2 Figure 3 MAIN Humans are capable of extracting relevant information when parsing sentences including the underlying hierarchical syntactic structures, and this remarkable hierarchical syntactic processing ability represents a milestone of human language evolution 1 – 7 . However, the neural underpinnings of syntactic processing are still controversial. One line of thoughts hold that the left inferior frontal gyrus (LIFG) subserves as a syntactic engine for hierarchical structure build-up 3,8–10 . Neuroimaging findings from basic merge 9 , 11 , 12 to processing the complex sentences with multi-level embeddings 13 – 17 , and from natural/jabberwocky sentence comprehension 4 , 7 , 15 , 18 , 19 to artificial grammar learning/processing 8,9,20–27 , consistently document the critical contribution of LIFG to hierarchical syntactic structure construction. On contrary, evidence from the lesion data (esp., the paragrammatism) is in support of the syntactic role of the left posterior temporal lobe (LpTL, including the left superior temporal gyrus/sulcus and the left middle temporal gyrus) 28 , 29 . Recent magnetoencephalography (MEG) findings also indicate that LpTL might be the key region for syntactic processes 30 – 32 . As two critical sites in the left-dominant fronto-temporal language network, the interaction between LIFG and LpTL is also of significant importance to elucidate their functional roles in syntactic processing. A prominent neurolingusitic model assumes that the LIFG might modulate the LpTL in (complex) sentence processing through the dorsal pathway in a top-down fashion, where hierarchical structures generated from LIFG will be integrated with semantics in LpTL 3 . An alternative lesion-mapping model suggests that LpTL generates the hierarchical structures and send them to LIFG for morphosyntactic processes and linearization 28 . Thus, the prerequisite of understanding the dynamic interactions between these two regions is to figure out which region(s) might be sensitive to hierarchical syntactic structure construction. Going beyond the correlational relationships provided by the neuroimaging results and the syntactic dysfunctions caused by the comparatively large lesion areas, non-invasive brain stimulation (NIBS) techniques such as transcranial magnetic stimulation (TMS) or ultrasound stimulation (TUS) could modulate the target brain regions in a specific fashion so as to provide the causal evidence of the syntactic role played by the candidate regions. Previous studies using TMS have reveale d the causal role of LIFG in syntactic processing 33 – 44 . In specific, the continuous theta-burst stimulation (cTBS) exhibits inhibitory effect of LIFG on sentence reading increasing with higher syntactic complexity of the sentences 44 . Moreover, adopting an explicit syntactic task which requires participants to label the syntactic categories of the testing structures, a multi-focal cTBS study on the language network composed of LIFG and LpTL showed a decrease in the processing stability 45 . However, direct comparison between the causal roles of LIFG and LpTL in syntactic processing is rare. A temporally specific TMS-Electroencephalogram (EEG) study 46 , using a two-word processing paradigm, demonstrates that the LIFG but not the LpTL is engaged in syntactic composition rather than the mere syntactic prediction. Schroën et al. 47 is among the first to integrate TMS with EEG to provide causal evidence for region-specific, time-critical processing-windows during auditory sentence comprehension, and specifies the complex interaction between LIFG and LpTL. But their work mainly focuses on semantic aspects: LpTL is initially involved in phonological and lexical processes in the early time-window, and then sends these information to LIFG in a bottom-up manner; LIFG, in turn, send the information to the LpTL where syntactic and semantic information integration may take place in a top-down manner. Nevertheless, since the syntactic complexity was not manipulated by the hierarchical embedding depths and non-adjacent dependency lengths (see also Gao et al. 45 ) in that study, direct causal evidence of the syntactic roles of LIFG and LpTL is still lacking. Notwithstanding the accumulating evidence indicates the causal role of the regions for syntactic processing, the application of TMS over the LIFG is criticized to cause significant pain. On the one hand, the TMS experiment might become unbearable; on the other hand, it is inevitable that participants can tell which is the real TMS session when the sham TMS is taken as a control condition. In contrast, TUS emerges as one of the state-of-the-art NIBS techniques, which is less painful and can be used for deep brain stimulation, potentially exciting neurons through a primarily mechanical mechanism mediated by ultrasound-induced opening of mechano-sensitive ion channels on the celluar membrane thereby perturbing neural activity 48 – 51 . TUS has been recently introduced from clinic treatments 52 , 52 , 53 to healthy cognitive operations 54 – 56 and to investigate semantic memory 57 . However, the use of tbTUS for the investigation in syntactic processing leaves a significant research gap. Therefore, the present study sets out to unravel the direct causal roles of LIFG and LpTL in syntactic processing during complex and simple sentence reading by utilizing TUS as a primary endeavor. Syntactic complexity effect is highlighted by the contrast between the complex sentences with relative clause embedded and the simpler semantically-matched coordinate sentences 44 , 58 (see also Fig. 1 A for materials). Several steps were taken to control various aspects of the TUS experiment. In addition to LIFG and LpTL, the vertex is chosen to be the active stimulation control site. Please refer to Fig. 1 for illustration of the experimental materials and procedures. In order to localize the peak coordinates (under the contrast of “complex sentence > simple sentence”) for modeling the TUS intensity we performed an additional fMRI study before the actual TUS experiment. Moreover, in order to reveal the causality trend while excluding the verbal working memory effect we conducted an additional TMS study (see Supplementary Information for details). ---- Insert Fig. 1 about here---- Based on the current discussion about the possible casual involvement of LIFG and LpTL we set out to identify the causal role of LIFG and/or LpTL in hierarchical syntactic processes using TUS (esp., tbTUS) with the goal to reveal consistent pattern efficaciously. We further complemented our conventional behavioral analyses with hierarchical drift-diffusion modelling (HDDM), to provide a principle way to test if TUS effects are best explained by altered evidence accumulation (drift rate) versus changes in response caution (boundary separation), or a combination of these parameters 59 , 60 . The present work not only seeks to deepen our understanding of the causal neural mechanisms underlying the unique human language faculty regarding the LIFG and LpTL of the critical cortical language circuit, but will also provide methodological references for TUS in future neurolinguistic research and clinical research especially related to sentence-level comprehension. RESULTS Detailed statistical outputs from the linear mixed models (LMMs) for reaction time (RT) and the general linear mixed models (GLMMs) for accuracy are summarized in Table 1 and 2. The distributions of the behavioral indices ( d’ , RT*, and CV ) for different sentence structures and different brain regions are illustrated in Figure 2. LIFG TUS caused prolonged RT. A significant main effect of Stimulation region ( F (2, 37.61) = 4.16, p = 0.023) for RT was detected. The post-hoc tests showed that LIFG TUS would prolong the RT than TUS on either LpTL ( b = 0.020, SE = 0.007, t = 2.86, p = 0.013) or vertex ( b = 0.014, SE = 0.006, z = 2.28, p = 0.068). The main effect of Structure type for RT was also significant: The RT of Complex sentence processing was much longer than that of simple sentence processing ( b = 0.049, SE = 0.003, t = 14.82, p < 0.001). No significant interaction could be found between Stimulation region and Structure type ( F = 0.52, p = 0.593) for RT, and there was no significant RT differences between LpTL TUS and vertex TUS ( z = 1.21, p = 0.685). LIFG TUS resulted in lower accuracy. As for accuracy, stimulation region showed a significant main effect ( χ² (2) = 12.86 , p = 0.002), in which LIFG TUS led to lower accuracy of sentence processing than LpTL TUS ( z = -2.961, p = 0.009) and vertex TUS ( z = -3.525, p = 0.001). Again, Structure type showed its significant main effect, in which the accuracy of complex sentence processing was lower than that of simple sentence processing ( b = -0.625, SE = 0.098, z = -6.41, p < 0.001). Moreover, Stimulation region and Structure type did not interacted significantly ( χ² (2) = 3.31 , p = 0.191) for accuracy. And sentence processing accuracy of LpTL TUS could not be differentiated from vertex TUS ( z = -1.13, p = 0.774). ---- Insert Table 1 and 2 about here---- LIFG TUS reduced the signal detection sensitivity. For the processing quality indices at the group-level analyses, Stimulation region showed a significant main effect for d’ in both Structure types ( F (1.70, 59.35) = 6.748, p = 0.004, η p 2 = 0.162), and the post-hoc tests further revealed that LIFG TUS significantly reduced the sentence processing sensitivity when compared with LpTL TUS ( t = -2.475, p = 0.055) and vertex TUS ( t = -3.355, p = 0.006). Moreover, a significant main effect of both Structure type was identified ( F (1, 35) = 15.129, p < 0.001, η p 2 = 0.302). There was no significant interaction between Stimulation region and Structure type ( F (1.74,60.73) = 2.245, p = 0.121, η p 2 = 0.060) for d’ , and the d’ differences between LpTL TUS and vertex TUS showed null results ( t = -0.701, p = 1.000). LIFG TUS deteriorated the integrative processing performance. The integrative RT * exhibited a consistent pattern. Stimulation region showed a significant main effect ( F (1.57, 54.83) = 11.772, p < 0.001, η p 2 = 0.252), where LIFG TUS elicited a much longer RT * than LpTL TUS ( t = 4.014, p < 0.001) and vertex TUS ( t = 3.440, p = 0.005). Structure type, again, showed its significant main effect ( F (1, 35) = 48.209, p < 0.001, η p 2 = 0.579). No significant interaction between Stimulation region and Structure type could be identified ( F (1.94,67.84) = 2.026, p = 0.141, η p 2 = 0.055), and there was no statistic differences between LpTL TUS and vertex TUS for RT * ( t = -0.721, p = 1.000). Processing stability was not significantly affected. Furthermore, processing stability, as reflected by the CV results, showed null results for Stimulation region ( F (1.80, 62.95) = 1.720, p = 0.190, η p 2 = 0.047), Structure type ( F (1, 35) = 0.106, p = 0.747, η p 2 = 0.003), and their interaction ( F (1.89, 66.22) = 0.260, p = 0.760, η p 2 = 0.007). ---- Insert Figure 2 about here---- LIFG TUS led to decreased drift rate during decision-making. Model comparison favored the model in which only drift rate varies across conditions, Model v, DIC = 34253) over the baseline model (Model 0, DIC = 35873), whereas and the full model in which all parameters vary (Model vat) failed to converge. This implies that the TUS effects are best captured by condition-dependent changes in drift rate. Under Model v, using the vertex as baseline, drift rate was lower for LIFG stimulation (mean = -0.14, 94% HDI [-0.19, -0.28]); whereas LpTL stimulation increased drift rate (mean = 0.09, 94% HDI [0.02, 0.15]). This suggests dissociable effects of LIFG and LpTL TUS on evidence accumulation during the decision-making process (Figure 3). However, the suspicious facilitatory effect of LpTL TUS could not result in shorter RT. We summarized convergence diagnostics for the primary drift-rate contrasts are summarized in Supplementary Information Table S5. ---- Insert Figure 3 about here---- Taken together, these results demonstrated that LIFG TUS rather than LpTL should be sensitive to the hierarchically-structured sequence (including both complex sentence and simple sentence) processing. We further decomposed the complex sentence condition into SRC and ORC subconditions for further analyses to evaluate the syntactic complexity effect, see Supporting Materials for more information. DISCUSSION The present study scrutinizes the causal role of the LIFG and LpTL in hierarchical syntactic processing through the TUS experiment. The observed significant inhibitory effects of LIFG TUS (but not LpTL TUS) on sentence processing quality and speed strengthen the causal role of LIFG rather than LpTL in hierarchical syntactic processes. To date, TUS, an emerging powerful NIBS technique, is initially introduced to the sentence-level syntactic processing by the present study. The 10–20 system for EEG has been proved to be an acceptable alternative solution for localizing the targets, when navigated TUS is not available 56 , 61 . In the present TUS study natural and concrete sentential meanings were bleached out by using geometric shapes and the unpredictable actions among these shapes to amplify the syntactic effects. Both the LMM and GLMM results indicated that LIFG plays a critical role in causing the hierarchical structure construction during sentence processing. At the first glance, LIFG TUS seems to be sensitive to all kinds of structure types. After all, all these structures are hierarchically constructed, and previous TMS experiments have differentiated the syntactic effects from the working memory capacity of the mere word lists 44 (see also the TMS findings reported in the Supplementary Information). Thus, LIFG might be sensitive to structured sequences no matter how syntactically complex they turn out to be 42,62 . However, previous neuroimaging results showed that the activation of LIFG increase for more complex syntactic combotorics in an artificial grammar processing experiment 8 , and the LIFG TMS showed that more complex the syntactic structure is, the more unstable syntactic processing would be 44,45 . Our exploratory analyses showed that LIFG TUS would diminish the d’ s and RT *s of complex sentence processing but not those of simple sentence processing (see Supplementary Information). Therefore, LIFG might not only respond to hierarchical syntactic processing, but also act sensitively to the syntactic complexity. Future studies may recruit more syntactic structure types to increase the variance of syntactic complexity so as to assess this hypothesis with sufficient statistic powers by performing regression modeling or representation similarity analysis (see also Gao et al. 45 ). Interestingly, the LIFG TUS showed its sound efficacy in processing quality indices rather than the index reflecting processing stability/auotmaticity at the group-level. In the recent TMS studies, the LIFG TMS or the LIFG-LpTL TMS is related to the change of the processing stability 44 , 45 (see also Supplementary Information for the TMS results). A plausible but tentative explanation might be due to the mechanical mechanism of TUS is different from the magnetic-electronic mechanism of TMS, which awaits to be further explored at the level of physic principles underlying these two NIBS techniques. Nevertheless, given the present study does not focus on the strict comparisons between TUS and TMS, we shall be concentrated on the replicablity of the findings demonstrating the causal role of LIFG in hierarchical syntactic processing even though the stimulation techniques as well as their protocols are distinct. Furthermore, on the basis of the results of HDDM, it is clear that LIFG TUS could perturb the drift rate of making the decision on “who is doing what to whom,” that is, the key process of sentence parsing. This means that LIFG TUS is able to affect the internal processing mechanisms of sentence comprehension. The unexpected facilitatory TUS effect of LpTL on the drift rate did not make the RTs shorter and awaits further explorations. It is also noteworthy that the inhibitory direction of LIFG tbTUS in the current study aligns with that of a recent deep brain stimulation study using the same protocol 49 , but note that the tbTUS protocol has previously been reported to induce facilitation 63 , 64 . Thus, the specific tasks for which tbTUS enhances or impairs performance, remain largely unclear yet. Although previous neuroimaging results detected the activation of LpTL in syntactic processing tasks 7 , 8 , 18 , 65 , 66 , its causal role in hierarchical syntactic processing did not receive sufficient empirical support. Moreover, brain lesions in LpTL might encompass broader areas, which could not be just attributed to the function of this region solely. The null effects of LpTL of the current study echoes well with the recent TMS studies that LpTL alone should not be causally involved in syntactic processing but rather in lexical-semantic processes 3 , 47 or late semantic syntactic integration processes. The LpTL should interact with LIFG to interfere with the syntactic processing when both regions are perturbed simultaneously 45 . Hence, in contrast to the claims of the syntactic role of LpTL (such as Matchin & Hickok 28 ), we are still lacking consistently-robust and thus convincing empirical evidence for the hierarchical syntactic processing role of LpTL (see also Uddén & Männel 67 ), also from the perspective of the brain-behavior causal relationships. To conclude, the present TUS experiment converged on the finding that the LIFG, but not the LpTL, plays a crucial causal role in hierarchical syntactic processing—though neural signal changes following stimulation of each region warrant further investigation in future studies. Moreover, as a novel NIBS technique in the field of sentence processing, TUS demonstrated efficacy in interfering with language-related regions to modulate associated behavioral performance. METHODS Participants Thirty-six healthy adult Chinese native speakers (Age: 21.4 ± 1.83 years; 24 females) underwent this experiment. They had normal or corrected-to-normal vision and were right-handed, with no history of psychiatric or neurological diseases. No participants failed or quit this experiment, so all the participants’ data were recruited for subsequent analyses. All participants gave signed informed consent before the experiment and received remuneration for participation. This study was approved by the ethics committee of the local university. Materials Following Thibault et al. 58 and Wu et al. 44 , The “complex sentences” in this study refer to sentences containing a center-embedding relative clause (RC). Such complex sentences are representative materials to reflect the complicated hierarchical nature of human languages and have been used as ideal experimental materials for decades (see the early works of Just et al. 65 and Stromswold et al. 68 ). We generated complex sentences with subject relative clause embedded (SRC) and complex sentences with object relative clause embedded (ORC). Each type of relative clause could appear at either the subject or the object position in the main clause. Simple coordinated sentences that semantically correspond to the complex sentences were also generated 58 . We utilized geometric shapes as nouns to minimize semantic interference and language experience 13 , 14 so as to somehow amplify the syntactic effects (e.g., “ 椭圆撞了弹了三角的方块 ”[The ellipse hit the squire that bounced the triangle]). Moreover, 96 complex sentences including 48 SRCs and 48 ORCs, and 96 simple sentences were assigned to each session to improve the statistic power. Furthermore, the occurrences of the single words and word pairs (such as a bigram composed of a noun and a verb or of two nouns/verbs) were carefully controlled so that participants were unable to make a response by a particular word or a word pair after reading each sentence. The experimental materials were summarized in Fig. 1 . Procedures The main procedures For each experiment, participants visited the lab for 3 times with an in-between interval of at least 7 days (see also Fig. 1 for illustration). The order of the stimulation sites/sessions were balanced across participants. In each session, they received offline TUS, and then were seated before the computer to complete the sentence processing task. Each trial started with a fixation (jittering from 0.5 to 2.5s), and then the testing sentence was presented for 4s. Participants were asked to select the correct answer to the two-option probe question regarding the thematic relations related to the test sentence within 3s. The stimuli were presented by E-prime 3.0 (Psychology Software Tools, Inc., Pittsburgh, PA, USA; https://support.pstnet.com ). The tbTUS protocol The present TUS experiment could be deemed as a pilot within-subject study which, for the first time, introduced (tb)TUS into the field of syntactic processing. The tbTUS protocol 63 , 64 was adopted and was defined as an 80 s train of 20 ms ultrasound pulse repeated every 200 ms. The TUS system (JL-TUS-200, Jiangxi Jielian Medical Equipment Co., Ltd., Nanchang, China) drove a transducer to generate ultrasonic waves at fundamental frequency of 0.5 MHz and focal depth of 25 mm. The acoustic intensity profile of the transducer has been measured in a previous study 69 with a fiber optic hydrophone in a custom-built water tank. The measurements indicated that the spatial peak-pulse-average intensity (I SPPA ) and the spatial-peak time-average intensity (I SPTA ) in free water were 9.67 W/cm 2 and 1.73 W/cm 2 . In addition, after transcranial transmission through a human skull, the I SPPA and I SPTA dropped by approximately fivefold compared to free water. LIFG, LpTL, and vertex were located based on international 10–20 system 56 , 61 , and we targeted LIFG, LpTL, and vertex via electroencephalography (EEG) electrode site F7, TP7, and Cz (Fig. 1 C for the spatial intensity distribution and pressure map for LIFG and LpTL, modeled by BabelBrain 70 ). For instance, F7 was mapped onto LIFG BA45/44 via the integration of EEG and MRI 71 , 72 . And these electrode sites were relatively close to the coordinates identified in our pilot fMRI study (see Supplementary Information). Conductive gel was applied between the transducer and head. Data analyses We employed the lme4 package and the lmerTest package in R 4.2.2 to contrast linear mixed effects models (LMMs) for response time (RT) of correctly-responded sequences, and accuracy will also be analyzed via comparing the generalized linear mixed-effects models (GLMMs) 73 – 76 . For the behavioral data of each experiment, the full model for RT or accuracy will be defined as: RT or Accuracy ~ Stimulation region × Structure type + (1 + Stimulation region × Structure type | Subject) + (1 + Stimulation region | Item) . To note, “ Stimulation region ” contails LIFG, LpTL, and vertex; “ Structure type ” incorporates the complex and the simple sentences. The Restricted Maximum Likelihood (REML) method will be used for parameter estimation. Fixed effects will be tested using a two-tailed t-test with the Kenward-Roger method for degrees of freedom estimation. Moreover, synthesized indices at the group-level, especially the indices reflecting the processing quality: d-prime ( d’ , reflecting sensitivity to the target) and accuracy-weighted RT ( RT *), and the index reflecting the processing stability: coefficient of variance ( CV , standing for processing stability or automaticity), will be also analyzed to evaluate the TUS efficacy in a relatively comprehensive manner via repeated measures analysis of variance (ANOVA) 44 . For d′ , we applied the signal detection theory formula: z (hit rate) − z (false-alarm rate). When hit or false-alarm rates equaled 0 or 1, a log-linear correction was applied by adding 0.5 to counts and 1 to totals to avoid infinite z -scores 77 . For RT *, the score was derived using the formula: RT * = RT (1 + 2 ER ), with ER representing the error rate 78 . For CV , this index was calculated as SD / mean RT 79 . We applied the DDM analysis using the HDDM toolbox (Hierarchical Drift-Diffusion Modeling 60 , version = 0.9.8RC). We fitted HDDM to the same trial-level dataset used in the behavioral mixed-effects models. We used HDDM’s accuracy-coding scheme by representing incorrect responses as negative and correct responses as positive RTs, so that both correct and incorrect trials contribute to model fitting. To reduce the influence of extreme RTs, trials with |RT| 4 s were excluded. In addition, we specified a contaminant process (p_outlier = 0.05) to down-weight residual outliers. We assumed the standard DDM. Starting point (z) was fixed at 0.5 (no response bias), accordingly, z was not estimated and was held constant across all models. We modelled the 3 (stimulation site) × 2 (sentence structure) design using HDDMRegressor with identity link functions. For parameters that were allowed to vary across experimental conditions, categorical regressors were used such that the vertex and simple sentences served as reference levels. This yields (1) stimulation-region contrasts referencing to the vertex stimulation and (2) complex - simple contrasts estimated separately within each stimulation region. We compared three nested models: a. Model 0: v, a, and t constant (no stimulation or complexity effects). b. Model v: v varies by stimulation × complexity ; a and t constant. c. Model vat: v, a, and t vary stimulation × complexity. Models were estimated using HDDM’s informative priors for v, a, and t. Subject-level parameters were estimated hierarchically. For each model, we ran 4 independent Markov chain Monte Carlo (MCMC) chains with 20,000 samples per chain. The first 5,000 samples of each chain were discarded as burn-in, and samples were thinned by a factor of 3.. Posterior summaries were computed with 94% highest-density intervals (HDIs). Convergence diagnostics (Gelman–Rubin R-hat) were extracted from the same summaries. We considered models whose R-hat < 1.1 as converging models. Candidate models with good convergence were compared using the deviance information criterion (DIC); lower DIC indicates a better trade-off between model fit and complexity. All inferential statements for parameter effects are based on posterior contrasts; an effect was considered credible when its 94% HDI excluded zero. Declarations DATA AND CODE AVAILABILITY STATEMENT Anonymized data and analysis code will be made available upon reasonable requests and collaborative agreement addressed to the coauthors. CONFLICT OF INTEREST STATEMENT The authors declare no competing interests. References Berwick RC, Friederici AD, Chomsky N, Bolhuis JJ (2013) Evolution, brain, and the nature of language. Trends Cogn Sci 17:89–98 Chomsky N (2014) The Minimalist Program. MIT Press Friederici AD (2017) Language in Our Brain: The Origins of a Uniquely Human Capacity. MIT Press Goucha T, Zaccarella E, Friederici AD (2017) A revival of Homo loquens as a builder of labeled structures: Neurocognitive considerations. Neurosci Biobehavioral Reviews 81:213–224 Everaert MB, Huybregts MA, Chomsky N, Berwick RC, Bolhuis JJ (2015) Structures, not strings: Linguistics as part of the cognitive sciences. Trends Cogn Sci 19:729–743 Miyagawa S, Berwick RC, Okanoya K (2013) The emergence of hierarchical structure in human language. Front Psychol 4:71 Pallier C, Devauchelle A-D, Dehaene S (2011) Cortical representation of the constituent structure of sentences. Proc. Natl. Acad. Sci. U.S.A. 108, 2522–2527 Chen L, Goucha T, Männel C, Friederici AD, Zaccarella E (2021) Hierarchical syntactic processing is beyond mere associating: Functional magnetic resonance imaging evidence from a novel artificial grammar. Hum Brain Mapp 42:3253–3268 Liu Y et al (2023) Exploring the neurobiology of Merge at a basic level: insights from a novel artificial grammar paradigm. Front Psychol 14:1151518 Zaccarella E, Schell M, Friederici AD (2017) Reviewing the functional basis of the syntactic Merge mechanism for language: A coordinate-based activation likelihood estimation meta-analysis. Neurosci Biobehavioral Reviews 80:646–656 Zaccarella E, Friederici AD (2015) Syntax in the brain. Brain Mapping: Encyclopedic Ref 461–468 Wu C, Zaccarella E, Friederici AD (2019) Universal neural basis of structure building evidenced by network modulations emerging from Broca’s area: The case of Chinese. Hum Brain Mapp 40:1705–1717 Indefrey P et al (2001) A neural correlate of syntactic encoding during speech production. Proc. Natl. Acad. Sci. U.S.A. 98, 5933–5936 Indefrey P, Hellwig F, Herzog H, Seitz RJ, Hagoort P (2004) Neural responses to the production and comprehension of syntax in identical utterances. Brain Lang 89:312–319 Makuuchi M, Bahlmann J, Anwander A, Friederici AD (2009) Segregating the core computational faculty of human language from working memory. Proc. Natl. Acad. Sci. U.S.A. 106, 8362–8367 Makuuchi M, Grodzinsky Y, Amunts K, Santi A, Friederici AD (2013) Processing noncanonical sentences in Broca’s region: reflections of movement distance and type. Cereb Cortex 23:694–702 Wang P et al (2021) Functional brain plasticity during L1 training on complex sentences: Changes in gamma-band oscillatory activity. Hum Brain Mapp 42:3858–3870 Chen L et al (2023) Frontotemporal effective connectivity revealed a language-general syntactic network for Mandarin Chinese. J Neurolinguistics 66:101127 Tyler LK et al (2010) Preserving Syntactic Processing across the Adult Life Span: The Modulation of the Frontotemporal Language System in the Context of Age-Related Atrophy. Cereb Cortex 20:352–364 Skeide MA, Brauer J, Friederici AD (2016) Brain functional and structural predictors of language performance. Cereb Cortex 26:2127–2139 Bahlmann J, Schubotz RI, Friederici AD (2008) Hierarchical artificial grammar processing engages Broca’s area. NeuroImage 42:525–534 Bahlmann J, Schubotz RI, Mueller JL, Koester D, Friederici AD (2009) Neural circuits of hierarchical visuo-spatial sequence processing. Brain Res 1298:161–170 Friederici AD, Bahlmann J, Heim S, Schubotz RI, Anwander A (2006) The brain differentiates human and non-human grammars: Functional localization and structural connectivity. Proc. Natl. Acad. Sci. U.S.A. 103, 2458–2463 Opitz B, Friederici AD (2003) Interactions of the hippocampal system and the prefrontal cortex in learning language-like rules. NeuroImage 19:1730–1737 Opitz B, Friederici AD (2004) Brain correlates of language learning: the neuronal dissociation of rule-based versus similarity-based learning. J Neurosci 24:8436–8440 Opitz B, Friederici AD (2007) Neural basis of processing sequential and hierarchical syntactic structures. Hum Brain Mapp 28:585–592 Ohta S, Fukui N, Sakai KL (2013) Syntactic computation in the human brain: the degree of merger as a key factor. PLoS ONE 8:e56230 Matchin W, Hickok G (2020) The cortical organization of syntax. Cereb Cortex 30:1481–1498 Matchin W et al (2024) Verbal working memory and syntactic comprehension segregate into the dorsal and ventral streams, respectively. Brain Commun 6:fcae449 Pylkkänen L (2019) The neural basis of combinatory syntax and semantics. Science 366:62–66 Law R, Pylkkänen L (2021) Lists with and without syntax: A new approach to measuring the neural processing of syntax. J Neurosci 41:2186–2196 Matar S, Dirani J, Marantz A, Pylkkänen L (2021) Left posterior temporal cortex is sensitive to syntax within conceptually matched Arabic expressions. Sci Rep 11:7181 Acheson DJ, Hagoort P (2013) Stimulating the Brain’s Language Network: Syntactic Ambiguity Resolution after TMS to the Inferior Frontal Gyrus and Middle Temporal Gyrus. J Cogn Neurosci 25:1664–1677 Carreiras M, Pattamadilok C, Meseguer E, Barber H, Devlin JT (2012) Broca’s area plays a causal role in morphosyntactic processing. Neuropsychologia 50:816–820 Coetzee JP et al (2022) Dissociating Language and Thought in Human Reasoning. Brain Sci 13:67 Kroczek LO, Gunter TC, Rysop AU, Friederici AD, Hartwigsen G (2019) Contributions of left frontal and temporal cortex to sentence comprehension: Evidence from simultaneous TMS-EEG. Cortex 115:86–98 Kuhnke P, Meyer L, Friederici AD, Hartwigsen G (2017) Left posterior inferior frontal gyrus is causally involved in reordering during sentence processing. NeuroImage 148:254–263 Maria-Korina S, THE ROLE OF BROCA’S AREA IN SYNTAX: A TMS STUDY ON WRITTEN GREEK LANGUAGE (2015). European Scientific Journal, ESJ https://eujournal.org/index.php/esj/article/view/6479 Meyer L, Elsner A, Turker S, Kuhnke P, Hartwigsen G (2018) Perturbation of left posterior prefrontal cortex modulates top-down processing in sentence comprehension. NeuroImage 181:598–604 Sakai KL, Noguchi Y, Takeuchi T, Watanabe E (2002) Selective Priming of Syntactic Processing by Event-Related Transcranial Magnetic Stimulation of Broca’s Area. Neuron 35:1177–1182 Uddén J et al (2008) The inferior frontal cortex in artificial syntax processing: An rTMS study. Brain Res 1224:69–78 Uddén J, Ingvar M, Hagoort P, Petersson KM (2017) Broca’s region: A causal role in implicit processing of grammars with crossed non-adjacent dependencies. Cognition 164:188–198 Van Der Burght CL, Numssen O, Schlaak B, Goucha T, Hartwigsen G (2023) Differential contributions of inferior frontal gyrus subregions to sentence processing guided by intonation. Hum Brain Mapp 44:585–598 Wu J et al (2024) Continuous theta-burst stimulation on the left posterior inferior frontal gyrus perturbs complex syntactic processing stability in mandarin Chinese. Neurobiol Lang 5:608–627 Gao C et al (2025) Continuous theta-burst stimulation demonstrates language-network-specific causal effects on syntactic processing. NeuroImage 306:121014 Maran M, Friederici AD, Zaccarella E (2022) Syntax through the looking glass: A review on two-word linguistic processing across behavioral, neuroimaging and neurostimulation studies. Neurosci Biobehavioral Reviews 142:104881 Schroën JAM et al (2023) Causal evidence for a coordinated temporal interplay within the language network. Proc. Natl. Acad. Sci. U.S.A. 120, e2306279120 Fomenko A, Neudorfer C, Dallapiazza RF, Kalia SK, Lozano AM (2018) Low-intensity ultrasound neuromodulation: An overview of mechanisms and emerging human applications. Brain Stimul 11:1209–1217 Darmani G et al (2025) Individualized non-invasive deep brain stimulation of the basal ganglia using transcranial ultrasound stimulation. Nat Commun 16:2693 O’Reilly MA (2024) Exploiting the mechanical effects of ultrasound for noninvasive therapy. Science 385:eadp7206 Sarica C et al (2022) Human Studies of Transcranial Ultrasound neuromodulation: A systematic review of effectiveness and safety. Brain Stimul 15:737–746 Monti MM, Schnakers C, Korb AS, Bystritsky A, Vespa PM (2016) Non-Invasive Ultrasonic Thalamic Stimulation in Disorders of Consciousness after Severe Brain Injury: A First-in-Man Report. Brain Stimulation: Basic Translational Clin Res Neuromodulation 9:940–941 Chen S-G et al (2020) Transcranial focused ultrasound pulsation suppresses pentylenetetrazol induced epilepsy in vivo . Brain Stimul 13:35–46 Yaakub S et al (2024) Non-invasive Ultrasonic Neuromodulation of the Human Nucleus Accumbens Impacts Reward Sensitivit. Preprint at https://doi.org/10.1101/2024.07.25.605068 Nakajima K et al (2022) A causal role of anterior prefrontal-putamen circuit for response inhibition revealed by transcranial ultrasound stimulation in humans. Cell Rep 40 Ziebell P et al (2023) Inhibition of midfrontal theta with transcranial ultrasound explains greater approach versus withdrawal behavior in humans. Brain Stimul 16:1278–1288 Jung J, Atkinson-Clement C, Kaiser M, Ralph MAL (2025) Transcranial focused ultrasound stimulation of the anterior temporal lobe enhances semantic memory by modulating brain morphology, neurochemistry and neural dynamics. 03.10.642483 Preprint at https://doi.org/10.1101/2025.03.10.642483 (2025) Thibault S et al (2021) Tool use and language share syntactic processes and neural patterns in the basal ganglia. Science 374:eabe0874 Ratcliff R, McKoon G (2008) The diffusion decision model: theory and data for two-choice decision tasks. Neural Comput 20:873–922 Wiecki TV, Sofer I, Frank MJ (2013) HDDM: Hierarchical Bayesian estimation of the drift-diffusion model in Python. Front neuroinformatics 7:14 Legon W et al (2014) Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans. Nat Neurosci 17:322–329 Uddén J, Bahlmann J (2012) A rostro-caudal gradient of structured sequence processing in the left inferior frontal gyrus. Philosophical Trans Royal Soc B: Biol Sci 367:2023–2032 Zeng K et al (2022) Induction of Human Motor Cortex Plasticity by Theta Burst Transcranial Ultrasound Stimulation. Ann Neurol 91:238–252 Zeng K et al (2024) Effects of different sonication parameters of theta burst transcranial ultrasound stimulation on human motor cortex. Brain Stimul 17:258–268 Just MA, Carpenter PA, Keller TA, Eddy WF, Thulborn KR (1996) Brain Activation Modulated by Sentence Comprehension. Science 274:114–116 Matchin W, Hammerly C, Lau E (2017) The role of the IFG and pSTS in syntactic prediction: Evidence from a parametric study of hierarchical structure in fMRI. Cortex 88:106–123 Uddén J, Männel C (2018) Artificial Grammar Learning and Its Neurobiology in Relation to Language Processing and Development. In: Rueschemeyer S-A, Gaskell MG (eds) The Oxford Handbook of Psycholinguistics. Oxford University Press, pp 755–783. doi: 10.1093/oxfordhb/9780198786825.013.33 . Stromswold K, Caplan D, Alpert N, Rauch S (1996) Localization of Syntactic Comprehension by Positron Emission Tomography. Brain Lang 52:452–473 Li Z et al (2025) Cross-species characterization of transcranial ultrasound propagation. Brain Stimul 18:164–172 Pichardo S, BabelBrain (2023) An Open-Source Application for Prospective Modeling of Transcranial Focused Ultrasound for Neuromodulation Applications. IEEE Trans Ultrason Ferroelectr Freq Control 70:587–599 Koessler L et al (2009) Automated cortical projection of EEG sensors: anatomical correlation via the international 10–10 system. NeuroImage 46:64–72 Scrivener CL, Reader AT (2022) Variability of EEG electrode positions and their underlying brain regions: visualizing gel artifacts from a simultaneous EEG-fMRI dataset. Brain Behav 12:e2476 Baayen RH, Davidson DJ, Bates DM (2008) Mixed-effects modeling with crossed random effects for subjects and items. J Mem Lang 59:390–412 Jaeger TF (2008) Categorical data analysis: Away from ANOVAs (transformation or not) and towards logit mixed models. J Mem Lang 59:434–446 Barr DJ (2013) Random effects structure for testing interactions in linear mixed-effects models. Front Psychol 4 Bates D, Mächler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models using lme4. J Stat Softw 67:1–48 Stanislaw H, Todorov N (1999) Calculation of signal detection theory measures. Behav Res Methods Instruments Computers 31:137–149 Lyons IM, Price GR, Vaessen A, Blomert L, Ansari D (2014) Numerical predictors of arithmetic success in grades 1–6. Dev Sci 17:714–726 Jeon H-A, Friederici AD (2015) Degree of automaticity and the prefrontal cortex. Trends Cogn Sci 19:244–250 Tables Table 1 Results of the Linear Mixed-Effects Model for TUS RT Fixed Effects Random Effects Subject Item b SE t p 95% CI Variance ICC Variance ICC Intercept 3.160 0.008 406.29 < 0.001 [3.144, 3.176] 0.002 0.127 0.001 0.070 LpTL - LIFG -0.014 0.006 -2.28 0.028 [-0.027, -0.002] 0.001 -- -- -- vertex - LpTL 0.006 0.005 -1.21 0.236 [-0.017, 0.004] 0.001 -- -- -- COM - SP -0.049 0.003 -14.82 < 0.001 [-0.055, -0.042] -- -- -- -- LpTL - LIFG : COM - SP 0.002 0.005 0.36 0.719 [-0.007, 0.011] -- -- -- -- vertex - LpTL : COM - SP -0.005 0.005 -1.01 0.313 [-0.014, 0.004] -- -- -- -- Note. Model: lmer ( RT ~ Stimulation region * Structure type + (1 + Stimulation region |subject) + (1 |item)). Table 2 Results of the Generalized Linear Mixed-Effects Model for TUS ACC Fixed Effects Random Effects Subject Item b SE z p 95% CI Variance ICC Variance ICC Intercept 2.319 0.124 18.62 < 0.001 [2.074, 2.562] 0.452 -- 1.017 -- LpTL - LIFG 0.266 0.09 2.96 0.003 [0.089, 0.441] 0.127 -- -- -- vertex - LpTL 0.082 0.072 1.13 0.258 [-0.059, 0.223] 0.015 -- -- -- COM - SP 0.625 0.098 6.41 < 0.001 [0.434, 0.816] -- -- -- -- LpTL - LIFG : COM - SP -0.198 0.119 -1.66 0.719 [-0.431, 0.035] -- -- -- -- vertex - LpTL : COM - SP 0.025 0.123 0.20 0.313 [-0.216, 0.266] -- -- -- -- Note. Model: glmer ( ACC ~ Stimulation region * Structure type + (1 + Stimulation region |subject) + (1 |item)). Additional Declarations There is NO Competing Interest. Supplementary Files TUSsupplementaryinformationforsubmission.docx Supplementary Information: Unraveling the Causal Role of Left Inferior Frontal Gyrus for Human Syntactic Processing by Transcranial Ultrasound Stimulation Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8717654","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":582632883,"identity":"039ec402-a90c-44c9-8cf8-cd1dda8e0199","order_by":0,"name":"harry-luyao Chen","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-0796-3323","institution":"Hong Kong Polytechnic University","correspondingAuthor":true,"prefix":"","firstName":"harry-luyao","middleName":"","lastName":"Chen","suffix":""},{"id":582632884,"identity":"6e4641c8-2ea6-4067-8f47-8bb793df4249","order_by":1,"name":"Guang Yang","email":"","orcid":"","institution":"Hong Kong Polytechnic University","correspondingAuthor":false,"prefix":"","firstName":"Guang","middleName":"","lastName":"Yang","suffix":""},{"id":582632885,"identity":"8e6b3a89-e6fd-45ce-a642-6601b0cba220","order_by":2,"name":"Xingfang Qu","email":"","orcid":"","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Xingfang","middleName":"","lastName":"Qu","suffix":""},{"id":582632886,"identity":"2213e715-4be5-4675-ae42-7cf2fdf26244","order_by":3,"name":"Dongwei Li","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Dongwei","middleName":"","lastName":"Li","suffix":""},{"id":582632887,"identity":"fffbedf7-d4b4-4c38-9c6b-0b60e9c2a627","order_by":4,"name":"Chenguang Zhao","email":"","orcid":"","institution":"Chinese Institute for Brain Research","correspondingAuthor":false,"prefix":"","firstName":"Chenguang","middleName":"","lastName":"Zhao","suffix":""},{"id":582632888,"identity":"a8a76769-b101-4e2e-b2cf-60bdeafa9d32","order_by":5,"name":"Junjie Wu","email":"","orcid":"","institution":"Tianjin Normal University","correspondingAuthor":false,"prefix":"","firstName":"Junjie","middleName":"","lastName":"Wu","suffix":""},{"id":582632889,"identity":"0b250ac5-1575-43ab-8b74-f049ee41e67d","order_by":6,"name":"Ke Zeng","email":"","orcid":"","institution":"Beijing Normal University","correspondingAuthor":false,"prefix":"","firstName":"Ke","middleName":"","lastName":"Zeng","suffix":""},{"id":582632890,"identity":"b366dd9c-a205-47e8-bfba-fc99e5f1f360","order_by":7,"name":"Yifei He","email":"","orcid":"","institution":"University of Marburg","correspondingAuthor":false,"prefix":"","firstName":"Yifei","middleName":"","lastName":"He","suffix":""},{"id":582632891,"identity":"61a5cd5c-4ef8-4010-a1dc-2be2229791e5","order_by":8,"name":"Gesa Hartwigsen","email":"","orcid":"https://orcid.org/0000-0002-8084-1330","institution":"MPI for Human Cognitive and Brain Sciences","correspondingAuthor":false,"prefix":"","firstName":"Gesa","middleName":"","lastName":"Hartwigsen","suffix":""},{"id":582632892,"identity":"00b4a857-5787-4798-8a87-d2ea24d11ed8","order_by":9,"name":"Robert Chen","email":"","orcid":"","institution":"University of Toronto","correspondingAuthor":false,"prefix":"","firstName":"Robert","middleName":"","lastName":"Chen","suffix":""},{"id":582632893,"identity":"72584af0-f5dc-4815-b021-d5bafaf70798","order_by":10,"name":"Angela D. Friederici","email":"","orcid":"","institution":"Max Planck Institute for Human Cognitive and Brain Sciences","correspondingAuthor":false,"prefix":"","firstName":"Angela","middleName":"D.","lastName":"Friederici","suffix":""}],"badges":[],"createdAt":"2026-01-28 07:46:38","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8717654/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8717654/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101507874,"identity":"3a42f931-04ae-4669-9a05-6884c87a526c","added_by":"auto","created_at":"2026-01-30 14:43:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":109577,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eExperimental materials and procedures.\u003c/strong\u003e\u003c/em\u003e (A) Illustration of syntactic manipulations across semantically natural and unnatural sequences. Sequences were constructed to contrast hierarchical structures differing in dependency length (subject-relative clause, SRC; object-relative clause, ORC; the trace [t] and the filler were co-indexed by the subscript “i”, and the key dependency was marked by the curve for each complex sentence type), and embedding depth (complex, COM; simple, SP). Examples were shown with both Chinese and English literal glosses and translations. (B) Experimental procedures. Participants received TUS over three sites—the left inferior frontal gyrus (LIFG), left posterior temporal lobe (LpTL), and vertex (i.e., the control site)—in a within-subjects design with sessions separated by ≥ 7 days. Each trial consisted of a jittered fixation period (0.5–2.5 s), sentence reading (4 s), and a probing question requiring response within 3s. (C) Spatial intensity distribution and pressure map for each region, modeled by BabelBrain. Group-level peak MNI coordinates are detected from the pilot fMRI study (see Supplementary Information for details).\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8717654/v1/dcd6daf178dd3280fcaacd66.png"},{"id":101507892,"identity":"3f1fca00-9250-4b03-b8db-8ffd24c5500a","added_by":"auto","created_at":"2026-01-30 14:43:10","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":102753,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003e\u003cstrong\u003eDistributions of significant main effects of synthesized indices across stimulation regions (A) and structure types (B).\u003c/strong\u003e\u003c/em\u003eEach panel shows violin plots representing individual participants’ data (dots), distribution density (violin shapes), and summary statistics (box plots indicating medians and interquartile ranges; whiskers showing data ranges excluding outliers). Horizontal brackets indicate pairwise contrasts between stimulation regions, with asterisks marking statistical significance levels derived from linear mixed-effects models. COM: complex sentence processing; SP: simple sentence processing. *\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05, **\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.01, ***\u003cem\u003ep\u003c/em\u003e\u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8717654/v1/1f11252f7ece9e0e96f0ab78.png"},{"id":101507798,"identity":"9820e8b0-1852-4ec8-8f85-cab9534d7086","added_by":"auto","created_at":"2026-01-30 14:42:55","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":29732,"visible":true,"origin":"","legend":"\u003cp\u003eHDDM posterior contrasts in drift rate. Posterior distributions of the stimulation effect on drift rate, expressed as contrasts relative to vertex stimulation (baseline; v = 0). Distributions shifted to positive values indicate higher drift rates (i.e., faster/more efficient evidence accumulation) compared with vertex stimulation. Shaded half-eye densities depict posterior samples; dots indicate posterior means; horizontal bars indicate 95% highest-density intervals (HDIs). The vertical dashed line marks the baseline. LIFG, left inferior frontal gyrus; LpTL, left posterior temporal lobe; HDDM, hierarchical drift-diffusion model; TUS, transcranial ultrasound stimulation.\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8717654/v1/5a9f9caafee323ee230a728e.png"},{"id":104398708,"identity":"b8f38c30-a6fe-4182-8ff3-082c4f853b06","added_by":"auto","created_at":"2026-03-11 12:03:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1100949,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8717654/v1/ed0cd32f-c2e6-4ef7-a998-df633799671e.pdf"},{"id":101507853,"identity":"0151d2a5-3347-4c5e-b905-5bd26c4f5795","added_by":"auto","created_at":"2026-01-30 14:43:01","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":2461384,"visible":true,"origin":"","legend":"Supplementary Information: Unraveling the Causal Role of Left Inferior Frontal Gyrus for Human Syntactic Processing by Transcranial Ultrasound Stimulation","description":"","filename":"TUSsupplementaryinformationforsubmission.docx","url":"https://assets-eu.researchsquare.com/files/rs-8717654/v1/d573188ece32e3830376662d.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Transcranial Ultrasound Stimulation Unravels the Causal Role of the Left Inferior Frontal Gyrus in Syntactic Processing","fulltext":[{"header":"MAIN","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003cp\u003eHumans are capable of extracting relevant information when parsing sentences including the underlying hierarchical syntactic structures, and this remarkable hierarchical syntactic processing ability represents a milestone of human language evolution\u003csup\u003e\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5 CR6\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. However, the neural underpinnings of syntactic processing are still controversial. One line of thoughts hold that the left inferior frontal gyrus (LIFG) subserves as a syntactic engine for hierarchical structure build-up\u003csup\u003e3,8\u0026ndash;10\u003c/sup\u003e. Neuroimaging findings from basic merge\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e to processing the complex sentences with multi-level embeddings\u003csup\u003e\u003cspan additionalcitationids=\"CR14 CR15 CR16\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, and from natural/jabberwocky sentence comprehension\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e to artificial grammar learning/processing\u003csup\u003e8,9,20\u0026ndash;27\u003c/sup\u003e, consistently document the critical contribution of LIFG to hierarchical syntactic structure construction. On contrary, evidence from the lesion data (esp., the paragrammatism) is in support of the syntactic role of the left posterior temporal lobe (LpTL, including the left superior temporal gyrus/sulcus and the left middle temporal gyrus) \u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. Recent magnetoencephalography (MEG) findings also indicate that LpTL might be the key region for syntactic processes\u003csup\u003e\u003cspan additionalcitationids=\"CR31\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. As two critical sites in the left-dominant fronto-temporal language network, the interaction between LIFG and LpTL is also of significant importance to elucidate their functional roles in syntactic processing. A prominent neurolingusitic model assumes that the LIFG might modulate the LpTL in (complex) sentence processing through the dorsal pathway in a top-down fashion, where hierarchical structures generated from LIFG will be integrated with semantics in LpTL\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. An alternative lesion-mapping model suggests that LpTL generates the hierarchical structures and send them to LIFG for morphosyntactic processes and linearization\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. Thus, the prerequisite of understanding the dynamic interactions between these two regions is to figure out which region(s) might be sensitive to hierarchical syntactic structure construction.\u003c/p\u003e \u003cp\u003e Going beyond the correlational relationships provided by the neuroimaging results and the syntactic dysfunctions caused by the comparatively large lesion areas, non-invasive brain stimulation (NIBS) techniques such as transcranial magnetic stimulation (TMS) or ultrasound stimulation (TUS) could modulate the target brain regions in a specific fashion so as to provide the causal evidence of the syntactic role played by the candidate regions. Previous studies using TMS have reveale\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003ed\u003c/span\u003e the causal role of LIFG in syntactic processing\u003csup\u003e\u003cspan additionalcitationids=\"CR34 CR35 CR36 CR37 CR38 CR39 CR40 CR41 CR42 CR43\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. In specific, the continuous theta-burst stimulation (cTBS) exhibits inhibitory effect of LIFG on sentence reading increasing with higher syntactic complexity of the sentences\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Moreover, adopting an explicit syntactic task which requires participants to label the syntactic categories of the testing structures, a multi-focal cTBS study on the language network composed of LIFG and LpTL showed a decrease in the processing stability\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. However, direct comparison between the causal roles of LIFG and LpTL in syntactic processing is rare. A temporally specific TMS-Electroencephalogram (EEG) study\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, using a two-word processing paradigm, demonstrates that the LIFG but not the LpTL is engaged in syntactic composition rather than the mere syntactic prediction. Schro\u0026euml;n et al.\u003csup\u003e47\u003c/sup\u003e is among the first to integrate TMS with EEG to provide causal evidence for region-specific, time-critical processing-windows during auditory sentence comprehension, and specifies the complex interaction between LIFG and LpTL. But their work mainly focuses on semantic aspects: LpTL is initially involved in phonological and lexical processes in the early time-window, and then sends these information to LIFG in a bottom-up manner; LIFG, in turn, send the information to the LpTL where syntactic and semantic information integration may take place in a top-down manner. Nevertheless, since the syntactic complexity was not manipulated by the hierarchical embedding depths and non-adjacent dependency lengths (see also Gao et al.\u003csup\u003e45\u003c/sup\u003e) in that study, direct causal evidence of the syntactic roles of LIFG and LpTL is still lacking.\u003c/p\u003e \u003cp\u003eNotwithstanding the accumulating evidence indicates the causal role of the regions for syntactic processing, the application of TMS over the LIFG is criticized to cause significant pain. On the one hand, the TMS experiment might become unbearable; on the other hand, it is inevitable that participants can tell which is the real TMS session when the sham TMS is taken as a control condition. In contrast, TUS emerges as one of the state-of-the-art NIBS techniques, which is less painful and can be used for deep brain stimulation, potentially exciting neurons through a primarily mechanical mechanism mediated by ultrasound-induced opening of mechano-sensitive ion channels on the celluar membrane thereby perturbing neural activity\u003csup\u003e\u003cspan additionalcitationids=\"CR49 CR50\" citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e. TUS has been recently introduced from clinic treatments\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e,\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e,\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e to healthy cognitive operations\u003csup\u003e\u003cspan additionalcitationids=\"CR55\" citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e and to investigate semantic memory\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e. However, the use of tbTUS for the investigation in syntactic processing leaves a significant research gap.\u003c/p\u003e \u003cp\u003eTherefore, the present study sets out to unravel the direct causal roles of LIFG and LpTL in syntactic processing during complex and simple sentence reading by utilizing TUS as a primary endeavor. Syntactic complexity effect is highlighted by the contrast between the complex sentences with relative clause embedded and the simpler semantically-matched coordinate sentences\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e,\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e(see also Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA for materials). Several steps were taken to control various aspects of the TUS experiment. In addition to LIFG and LpTL, the vertex is chosen to be the active stimulation control site. Please refer to Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e for illustration of the experimental materials and procedures. In order to localize the peak coordinates (under the contrast of \u0026ldquo;complex sentence\u0026thinsp;\u0026gt;\u0026thinsp;simple sentence\u0026rdquo;) for modeling the TUS intensity we performed an additional fMRI study before the actual TUS experiment. Moreover, in order to reveal the causality trend while excluding the verbal working memory effect we conducted an additional TMS study (see Supplementary Information for details).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e---- Insert Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e about here----\u003c/p\u003e \u003cp\u003eBased on the current discussion about the possible casual involvement of LIFG and LpTL we set out to identify the causal role of LIFG and/or LpTL in hierarchical syntactic processes using TUS (esp., tbTUS) with the goal to reveal consistent pattern efficaciously. We further complemented our conventional behavioral analyses with hierarchical drift-diffusion modelling (HDDM), to provide a principle way to test if TUS effects are best explained by altered evidence accumulation (drift rate) versus changes in response caution (boundary separation), or a combination of these parameters\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e,\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e. The present work not only seeks to deepen our understanding of the causal neural mechanisms underlying the unique human language faculty regarding the LIFG and LpTL of the critical cortical language circuit, but will also provide methodological references for TUS in future neurolinguistic research and clinical research especially related to sentence-level comprehension.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cp\u003eDetailed statistical outputs from the linear mixed models (LMMs) for reaction time (RT) and the general linear mixed models (GLMMs) for accuracy are summarized in Table 1 and 2. The distributions of the behavioral indices (\u003cem\u003ed\u0026rsquo;\u003c/em\u003e, \u003cem\u003eRT*,\u0026nbsp;\u003c/em\u003eand\u003cem\u003e\u0026nbsp;CV\u003c/em\u003e)\u0026nbsp;for different\u0026nbsp;sentence\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003estructures\u0026nbsp;and different brain regions\u0026nbsp;are illustrated in Figure 2.\u003c/p\u003e\n\u003cp\u003eLIFG TUS caused prolonged RT. A significant main effect of Stimulation region (\u003cem\u003eF\u003c/em\u003e(2, 37.61)\u003cem\u003e\u0026nbsp;\u003c/em\u003e= 4.16, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.023) for RT was detected. The post-hoc tests showed that LIFG TUS would prolong the RT than TUS on either LpTL (\u003cem\u003eb\u003c/em\u003e = 0.020, \u003cem\u003eSE\u003c/em\u003e = 0.007, \u003cem\u003et\u0026nbsp;\u003c/em\u003e= 2.86, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.013) or vertex (\u003cem\u003eb\u003c/em\u003e = 0.014, \u003cem\u003eSE\u003c/em\u003e = 0.006, \u003cem\u003ez\u0026nbsp;\u003c/em\u003e= 2.28, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.068). The main effect of Structure type for RT was also significant: The RT of Complex sentence processing was much longer than that of simple sentence processing (\u003cem\u003eb\u003c/em\u003e = 0.049, \u003cem\u003eSE\u003c/em\u003e = 0.003, \u003cem\u003et\u0026nbsp;\u003c/em\u003e= 14.82, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; 0.001). No significant interaction could be found between Stimulation region and Structure type (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e= 0.52, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.593) for RT, and there was no significant RT differences between LpTL TUS and vertex TUS (\u003cem\u003ez\u0026nbsp;\u003c/em\u003e= 1.21, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.685).\u003c/p\u003e\n\u003cp\u003eLIFG TUS resulted in lower accuracy. As for accuracy, stimulation region showed a significant main effect (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e(2) = 12.86\u003cem\u003e, p\u003c/em\u003e = 0.002), in which LIFG TUS led to lower accuracy of sentence processing than LpTL TUS (\u003cem\u003ez\u0026nbsp;\u003c/em\u003e= -2.961, \u003cem\u003ep =\u003c/em\u003e 0.009) and vertex TUS (\u003cem\u003ez\u0026nbsp;\u003c/em\u003e= -3.525, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.001). Again, Structure type showed its significant main effect, in which the accuracy of complex sentence processing was lower than that of simple sentence processing (\u003cem\u003eb\u003c/em\u003e = -0.625, \u003cem\u003eSE\u003c/em\u003e = 0.098, \u003cem\u003ez\u0026nbsp;\u003c/em\u003e= -6.41, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e\u0026lt; 0.001). Moreover, Stimulation region and Structure type did not interacted significantly (\u003cem\u003e\u0026chi;\u0026sup2;\u003c/em\u003e(2) = 3.31\u003cem\u003e, p\u003c/em\u003e = 0.191) for accuracy. And sentence processing accuracy of LpTL TUS could not be differentiated from vertex TUS (\u003cem\u003ez\u0026nbsp;\u003c/em\u003e= -1.13, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.774).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e---- Insert Table 1 and 2 about here----\u003c/p\u003e\n\u003cp\u003eLIFG TUS reduced the signal detection sensitivity. For the processing quality indices at the group-level analyses, Stimulation region showed a significant main effect for \u003cem\u003ed\u0026rsquo;\u003c/em\u003e in both Structure types (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1.70, 59.35) = 6.748, \u003cem\u003ep\u003c/em\u003e = 0.004,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.162), and the post-hoc tests further revealed that LIFG TUS significantly reduced the sentence processing sensitivity when compared with LpTL TUS (\u003cem\u003et\u0026nbsp;\u003c/em\u003e= -2.475, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.055) and vertex TUS (\u003cem\u003et\u003c/em\u003e = -3.355, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.006). Moreover, a significant main effect of both Structure type was identified (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1, 35) = 15.129, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.302). There was no significant interaction between Stimulation region and Structure type (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1.74,60.73)\u0026nbsp;=\u0026nbsp;2.245,\u0026nbsp;\u003cem\u003ep\u003c/em\u003e =\u0026nbsp;0.121,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.060) for \u003cem\u003ed\u0026rsquo;\u003c/em\u003e, and the \u003cem\u003ed\u0026rsquo;\u003c/em\u003e differences between LpTL TUS and vertex TUS showed null results (\u003cem\u003et\u0026nbsp;\u003c/em\u003e=\u0026nbsp;-0.701,\u0026nbsp;\u003cem\u003ep\u003c/em\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e= 1.000).\u003c/p\u003e\n\u003cp\u003eLIFG TUS deteriorated the integrative processing performance. The integrative \u003cem\u003eRT\u003c/em\u003e* exhibited a consistent pattern. Stimulation region showed a significant main effect (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1.57, 54.83) = 11.772, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.252), where LIFG TUS elicited a much longer \u003cem\u003eRT\u003c/em\u003e* than LpTL TUS (\u003cem\u003et\u0026nbsp;\u003c/em\u003e= 4.014, \u003cem\u003ep \u0026lt;\u003c/em\u003e 0.001) and vertex TUS (\u003cem\u003et\u0026nbsp;\u003c/em\u003e= 3.440, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 0.005). Structure type, again, showed its significant main effect (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1, 35) = 48.209, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.579). No significant interaction between Stimulation region and Structure type could be identified (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1.94,67.84)\u0026nbsp;=\u0026nbsp;2.026,\u0026nbsp;\u003cem\u003ep\u003c/em\u003e =\u0026nbsp;0.141,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.055), and there was no statistic differences between LpTL TUS and vertex TUS for \u003cem\u003eRT\u003c/em\u003e* (\u003cem\u003et\u0026nbsp;\u003c/em\u003e=\u0026nbsp;-0.721, \u003cem\u003ep\u0026nbsp;\u003c/em\u003e= 1.000).\u003c/p\u003e\n\u003cp\u003eProcessing stability was not significantly affected. Furthermore, processing stability, as reflected by the CV results, showed null results for Stimulation region (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1.80, 62.95) = 1.720, \u003cem\u003ep\u003c/em\u003e = 0.190,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.047), Structure type (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1, 35) = 0.106, \u003cem\u003ep\u003c/em\u003e = 0.747,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.003), and their interaction (\u003cem\u003eF\u0026nbsp;\u003c/em\u003e(1.89, 66.22) = 0.260, \u003cem\u003ep\u003c/em\u003e = 0.760,\u0026nbsp;\u003cem\u003e\u0026eta;\u003csub\u003ep\u003c/sub\u003e\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e = 0.007).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e---- Insert Figure 2 about here----\u003c/p\u003e\n\u003cp\u003eLIFG TUS led to decreased drift rate during decision-making. Model comparison favored the model in which only drift rate varies across conditions, Model v, DIC = 34253) over the baseline model (Model 0, DIC = 35873), whereas and the full model in which all parameters vary (Model vat) failed to converge. This implies that the TUS effects are best captured by condition-dependent changes in drift rate. Under Model v, using the vertex as baseline, drift rate was lower for LIFG stimulation (mean = -0.14, 94% HDI [-0.19, -0.28]); whereas LpTL stimulation increased drift rate (mean = 0.09, 94% HDI [0.02, 0.15]). This suggests dissociable effects of LIFG and LpTL TUS on evidence accumulation during the decision-making process (Figure 3). However, the suspicious facilitatory effect of LpTL TUS could not result in shorter RT. We summarized convergence diagnostics for the primary drift-rate contrasts are summarized in \u003cem\u003eSupplementary Information\u003c/em\u003e Table S5.\u003c/p\u003e\n\u003cp\u003e---- Insert Figure 3 about here----\u003c/p\u003e\n\u003cp\u003eTaken together, these results demonstrated that LIFG TUS rather than LpTL should be sensitive to the hierarchically-structured sequence (including both complex sentence and simple sentence) processing. We further decomposed the complex sentence condition into SRC and ORC subconditions for further analyses to evaluate the syntactic complexity effect, see Supporting Materials for more information.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe present study scrutinizes the causal role of the LIFG and LpTL in hierarchical syntactic processing through the TUS experiment. The observed significant inhibitory effects of LIFG TUS (but not LpTL TUS) on sentence processing quality and speed strengthen the causal role of LIFG rather than LpTL in hierarchical syntactic processes.\u003c/p\u003e \u003cp\u003eTo date, TUS, an emerging powerful NIBS technique, is initially introduced to the sentence-level syntactic processing by the present study. The 10\u0026ndash;20 system for EEG has been proved to be an acceptable alternative solution for localizing the targets, when navigated TUS is not available\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e,\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e. In the present TUS study natural and concrete sentential meanings were bleached out by using geometric shapes and the unpredictable actions among these shapes to amplify the syntactic effects. Both the LMM and GLMM results indicated that LIFG plays a critical role in causing the hierarchical structure construction during sentence processing. At the first glance, LIFG TUS seems to be sensitive to all kinds of structure types. After all, all these structures are hierarchically constructed, and previous TMS experiments have differentiated the syntactic effects from the working memory capacity of the mere word lists\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e (see also the TMS findings reported in the Supplementary Information). Thus, LIFG might be sensitive to structured sequences no matter how syntactically complex they turn out to be\u003csup\u003e42,62\u003c/sup\u003e. However, previous neuroimaging results showed that the activation of LIFG increase for more complex syntactic combotorics in an artificial grammar processing experiment\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, and the LIFG TMS showed that more complex the syntactic structure is, the more unstable syntactic processing would be\u003csup\u003e44,45\u003c/sup\u003e. Our exploratory analyses showed that LIFG TUS would diminish the \u003cem\u003ed\u0026rsquo;\u003c/em\u003es and \u003cem\u003eRT\u003c/em\u003e*s of complex sentence processing but not those of simple sentence processing (see Supplementary Information). Therefore, LIFG might not only respond to hierarchical syntactic processing, but also act sensitively to the syntactic complexity. Future studies may recruit more syntactic structure types to increase the variance of syntactic complexity so as to assess this hypothesis with sufficient statistic powers by performing regression modeling or representation similarity analysis (see also Gao et al.\u003csup\u003e45\u003c/sup\u003e).\u003c/p\u003e \u003cp\u003eInterestingly, the LIFG TUS showed its sound efficacy in processing quality indices rather than the index reflecting processing stability/auotmaticity at the group-level. In the recent TMS studies, the LIFG TMS or the LIFG-LpTL TMS is related to the change of the processing stability\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e,\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e (see also Supplementary Information for the TMS results). A plausible but tentative explanation might be due to the mechanical mechanism of TUS is different from the magnetic-electronic mechanism of TMS, which awaits to be further explored at the level of physic principles underlying these two NIBS techniques. Nevertheless, given the present study does not focus on the strict comparisons between TUS and TMS, we shall be concentrated on the replicablity of the findings demonstrating the causal role of LIFG in hierarchical syntactic processing even though the stimulation techniques as well as their protocols are distinct. Furthermore, on the basis of the results of HDDM, it is clear that LIFG TUS could perturb the drift rate of making the decision on \u0026ldquo;who is doing what to whom,\u0026rdquo; that is, the key process of sentence parsing. This means that LIFG TUS is able to affect the internal processing mechanisms of sentence comprehension. The unexpected facilitatory TUS effect of LpTL on the drift rate did not make the RTs shorter and awaits further explorations. It is also noteworthy that the inhibitory direction of LIFG tbTUS in the current study aligns with that of a recent deep brain stimulation study using the same protocol\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e, but note that the tbTUS protocol has previously been reported to induce facilitation\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e,\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e. Thus, the specific tasks for which tbTUS enhances or impairs performance, remain largely unclear yet.\u003c/p\u003e \u003cp\u003eAlthough previous neuroimaging results detected the activation of LpTL in syntactic processing tasks\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e,\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e, its causal role in hierarchical syntactic processing did not receive sufficient empirical support. Moreover, brain lesions in LpTL might encompass broader areas, which could not be just attributed to the function of this region solely. The null effects of LpTL of the current study echoes well with the recent TMS studies that LpTL alone should not be causally involved in syntactic processing but rather in lexical-semantic processes\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e or late semantic syntactic integration processes. The LpTL should interact with LIFG to interfere with the syntactic processing when both regions are perturbed simultaneously\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Hence, in contrast to the claims of the syntactic role of LpTL (such as Matchin \u0026amp; Hickok\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e), we are still lacking consistently-robust and thus convincing empirical evidence for the hierarchical syntactic processing role of LpTL (see also Udd\u0026eacute;n \u0026amp; M\u0026auml;nnel\u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e), also from the perspective of the brain-behavior causal relationships.\u003c/p\u003e \u003cp\u003eTo conclude, the present TUS experiment converged on the finding that the LIFG, but not the LpTL, plays a crucial causal role in hierarchical syntactic processing\u0026mdash;though neural signal changes following stimulation of each region warrant further investigation in future studies. Moreover, as a novel NIBS technique in the field of sentence processing, TUS demonstrated efficacy in interfering with language-related regions to modulate associated behavioral performance.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eThirty-six healthy adult Chinese native speakers (Age: 21.4\u0026thinsp;\u0026plusmn;\u0026thinsp;1.83 years; 24 females) underwent this experiment. They had normal or corrected-to-normal vision and were right-handed, with no history of psychiatric or neurological diseases. No participants failed or quit this experiment, so all the participants\u0026rsquo; data were recruited for subsequent analyses. All participants gave signed informed consent before the experiment and received remuneration for participation. This study was approved by the ethics committee of the local university.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eMaterials\u003c/h3\u003e\n\u003cp\u003eFollowing Thibault et al.\u003csup\u003e58\u003c/sup\u003e and Wu et al.\u003csup\u003e44\u003c/sup\u003e, The \u0026ldquo;complex sentences\u0026rdquo; in this study refer to sentences containing a center-embedding relative clause (RC). Such complex sentences are representative materials to reflect the complicated hierarchical nature of human languages and have been used as ideal experimental materials for decades (see the early works of Just et al.\u003csup\u003e65\u003c/sup\u003e and Stromswold et al.\u003csup\u003e68\u003c/sup\u003e). We generated complex sentences with subject relative clause embedded (SRC) and complex sentences with object relative clause embedded (ORC). Each type of relative clause could appear at either the subject or the object position in the main clause. Simple coordinated sentences that semantically correspond to the complex sentences were also generated\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. We utilized geometric shapes as nouns to minimize semantic interference and language experience\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e so as to somehow amplify the syntactic effects (e.g., \u0026ldquo;\u003cem\u003e椭圆撞了弹了三角的方块\u003c/em\u003e\u0026rdquo;[The ellipse hit the squire that bounced the triangle]). Moreover, 96 complex sentences including 48 SRCs and 48 ORCs, and 96 simple sentences were assigned to each session to improve the statistic power. Furthermore, the occurrences of the single words and word pairs (such as a bigram composed of a noun and a verb or of two nouns/verbs) were carefully controlled so that participants were unable to make a response by a particular word or a word pair after reading each sentence. The experimental materials were summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eProcedures\u003c/h2\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003eThe main procedures\u003c/h2\u003e \u003cp\u003eFor each experiment, participants visited the lab for 3 times with an in-between interval of at least 7 days (see also Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e for illustration). The order of the stimulation sites/sessions were balanced across participants. In each session, they received offline TUS, and then were seated before the computer to complete the sentence processing task. Each trial started with a fixation (jittering from 0.5 to 2.5s), and then the testing sentence was presented for 4s. Participants were asked to select the correct answer to the two-option probe question regarding the thematic relations related to the test sentence within 3s. The stimuli were presented by E-prime 3.0 (Psychology Software Tools, Inc., Pittsburgh, PA, USA; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://support.pstnet.com\u003c/span\u003e\u003cspan address=\"https://support.pstnet.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003ch3\u003eThe tbTUS protocol\u003c/h3\u003e\n\u003cp\u003eThe present TUS experiment could be deemed as a pilot within-subject study which, for the first time, introduced (tb)TUS into the field of syntactic processing. The tbTUS protocol\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e,\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e\u003c/sup\u003e was adopted and was defined as an 80 s train of 20 ms ultrasound pulse repeated every 200 ms. The TUS system (JL-TUS-200, Jiangxi Jielian Medical Equipment Co., Ltd., Nanchang, China) drove a transducer to generate ultrasonic waves at fundamental frequency of 0.5 MHz and focal depth of 25 mm. The acoustic intensity profile of the transducer has been measured in a previous study\u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e with a fiber optic hydrophone in a custom-built water tank. The measurements indicated that the spatial peak-pulse-average intensity (I\u003csub\u003eSPPA\u003c/sub\u003e) and the spatial-peak time-average intensity (I\u003csub\u003eSPTA\u003c/sub\u003e) in free water were 9.67 W/cm\u003csup\u003e2\u003c/sup\u003e and 1.73 W/cm\u003csup\u003e2\u003c/sup\u003e. In addition, after transcranial transmission through a human skull, the I\u003csub\u003eSPPA\u003c/sub\u003e and I\u003csub\u003eSPTA\u003c/sub\u003e dropped by approximately fivefold compared to free water. LIFG, LpTL, and vertex were located based on international 10\u0026ndash;20 system \u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e,\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e\u003c/sup\u003e, and we targeted LIFG, LpTL, and vertex via electroencephalography (EEG) electrode site F7, TP7, and Cz (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC for the spatial intensity distribution and pressure map for LIFG and LpTL, modeled by BabelBrain\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e). For instance, F7 was mapped onto LIFG BA45/44 via the integration of EEG and MRI\u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e,\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e. And these electrode sites were relatively close to the coordinates identified in our pilot fMRI study (see Supplementary Information). Conductive gel was applied between the transducer and head.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eData analyses\u003c/h2\u003e \u003cp\u003eWe employed the lme4 package and the lmerTest package in R 4.2.2 to contrast linear mixed effects models (LMMs) for response time (RT) of correctly-responded sequences, and accuracy will also be analyzed via comparing the generalized linear mixed-effects models (GLMMs) \u003csup\u003e\u003cspan additionalcitationids=\"CR74 CR75\" citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e. For the behavioral data of each experiment, the full model for RT or accuracy will be defined as:\u003c/p\u003e \u003cp\u003e \u003cem\u003eRT or Accuracy\u0026thinsp;~\u0026thinsp;Stimulation region \u0026times; Structure type + (1\u0026thinsp;+\u0026thinsp;Stimulation region \u0026times; Structure type | Subject) + (1\u0026thinsp;+\u0026thinsp;Stimulation region | Item)\u003c/em\u003e.\u003c/p\u003e \u003cp\u003eTo note, \u0026ldquo;\u003cem\u003eStimulation region\u003c/em\u003e\u0026rdquo; contails LIFG, LpTL, and vertex; \u0026ldquo;\u003cem\u003eStructure type\u003c/em\u003e\u0026rdquo; incorporates the complex and the simple sentences. The Restricted Maximum Likelihood (REML) method will be used for parameter estimation. Fixed effects will be tested using a two-tailed t-test with the Kenward-Roger method for degrees of freedom estimation.\u003c/p\u003e \u003cp\u003eMoreover, synthesized indices at the group-level, especially the indices reflecting the processing quality: d-prime (\u003cem\u003ed\u0026rsquo;\u003c/em\u003e, reflecting sensitivity to the target) and accuracy-weighted \u003cem\u003eRT\u003c/em\u003e (\u003cem\u003eRT\u003c/em\u003e*), and the index reflecting the processing stability: coefficient of variance (\u003cem\u003eCV\u003c/em\u003e, standing for processing stability or automaticity), will be also analyzed to evaluate the TUS efficacy in a relatively comprehensive manner via repeated measures analysis of variance (ANOVA)\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. For \u003cem\u003ed\u0026prime;\u003c/em\u003e, we applied the signal detection theory formula: \u003cem\u003ez\u003c/em\u003e(hit rate)\u0026thinsp;\u0026minus;\u0026thinsp;\u003cem\u003ez\u003c/em\u003e(false-alarm rate). When hit or false-alarm rates equaled 0 or 1, a log-linear correction was applied by adding 0.5 to counts and 1 to totals to avoid infinite \u003cem\u003ez\u003c/em\u003e-scores\u003csup\u003e\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e\u003c/sup\u003e. For \u003cem\u003eRT\u003c/em\u003e*, the score was derived using the formula: \u003cem\u003eRT\u003c/em\u003e* = \u003cem\u003eRT\u003c/em\u003e(1\u0026thinsp;+\u0026thinsp;2\u003cem\u003eER\u003c/em\u003e), with \u003cem\u003eER\u003c/em\u003e representing the error rate\u003csup\u003e\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e\u003c/sup\u003e. For \u003cem\u003eCV\u003c/em\u003e, this index was calculated as \u003cem\u003eSD\u003c/em\u003e / mean \u003cem\u003eRT\u003c/em\u003e\u003csup\u003e79\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eWe applied the DDM analysis using the HDDM toolbox (Hierarchical Drift-Diffusion Modeling\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e, version\u0026thinsp;=\u0026thinsp;0.9.8RC). We fitted HDDM to the same trial-level dataset used in the behavioral mixed-effects models. We used HDDM\u0026rsquo;s accuracy-coding scheme by representing incorrect responses as negative and correct responses as positive RTs, so that both correct and incorrect trials contribute to model fitting. To reduce the influence of extreme RTs, trials with |RT| \u0026lt; 0.15 s or |RT| \u0026gt; 4 s were excluded. In addition, we specified a contaminant process (p_outlier\u0026thinsp;=\u0026thinsp;0.05) to down-weight residual outliers. We assumed the standard DDM. Starting point (z) was fixed at 0.5 (no response bias), accordingly, z was not estimated and was held constant across all models. We modelled the 3 (stimulation site) \u0026times; 2 (sentence structure) design using HDDMRegressor with identity link functions. For parameters that were allowed to vary across experimental conditions, categorical regressors were used such that the vertex and simple sentences served as reference levels. This yields (1) stimulation-region contrasts referencing to the vertex stimulation and (2) complex - simple contrasts estimated separately within each stimulation region. We compared three nested models:\u003c/p\u003e \u003cp\u003ea. Model 0: v, a, and t constant (no stimulation or complexity effects).\u003c/p\u003e\u003cp\u003eb. Model v: v varies by stimulation \u0026times; complexity ; a and t constant.\u003c/p\u003e\u003cp\u003ec. Model vat: v, a, and t vary stimulation \u0026times; complexity.\u003c/p\u003e \u003cp\u003eModels were estimated using HDDM\u0026rsquo;s informative priors for v, a, and t. Subject-level parameters were estimated hierarchically. For each model, we ran 4 independent Markov chain Monte Carlo (MCMC) chains with 20,000 samples per chain. The first 5,000 samples of each chain were discarded as burn-in, and samples were thinned by a factor of 3.. Posterior summaries were computed with 94% highest-density intervals (HDIs). Convergence diagnostics (Gelman\u0026ndash;Rubin R-hat) were extracted from the same summaries. We considered models whose R-hat\u0026thinsp;\u0026lt;\u0026thinsp;1.1 as converging models. Candidate models with good convergence were compared using the deviance information criterion (DIC); lower DIC indicates a better trade-off between model fit and complexity. All inferential statements for parameter effects are based on posterior contrasts; an effect was considered credible when its 94% HDI excluded zero.\u003c/p\u003e \u003c/div\u003e "},{"header":"Declarations","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eDATA AND CODE AVAILABILITY STATEMENT\u003c/h2\u003e \u003cp\u003eAnonymized data and analysis code will be made available upon reasonable requests and collaborative agreement addressed to the coauthors.\u003c/p\u003e \u003c/div\u003e\u003ch2\u003eCONFLICT OF INTEREST STATEMENT\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e "},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBerwick RC, Friederici AD, Chomsky N, Bolhuis JJ (2013) Evolution, brain, and the nature of language. Trends Cogn Sci 17:89\u0026ndash;98\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChomsky N (2014) The Minimalist Program. MIT Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriederici AD (2017) Language in Our Brain: The Origins of a Uniquely Human Capacity. MIT Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoucha T, Zaccarella E, Friederici AD (2017) A revival of Homo loquens as a builder of labeled structures: Neurocognitive considerations. Neurosci Biobehavioral Reviews 81:213\u0026ndash;224\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEveraert MB, Huybregts MA, Chomsky N, Berwick RC, Bolhuis JJ (2015) Structures, not strings: Linguistics as part of the cognitive sciences. Trends Cogn Sci 19:729\u0026ndash;743\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiyagawa S, Berwick RC, Okanoya K (2013) The emergence of hierarchical structure in human language. Front Psychol 4:71\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePallier C, Devauchelle A-D, Dehaene S (2011) Cortical representation of the constituent structure of sentences. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 108, 2522\u0026ndash;2527\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen L, Goucha T, M\u0026auml;nnel C, Friederici AD, Zaccarella E (2021) Hierarchical syntactic processing is beyond mere associating: Functional magnetic resonance imaging evidence from a novel artificial grammar. Hum Brain Mapp 42:3253\u0026ndash;3268\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y et al (2023) Exploring the neurobiology of Merge at a basic level: insights from a novel artificial grammar paradigm. Front Psychol 14:1151518\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZaccarella E, Schell M, Friederici AD (2017) Reviewing the functional basis of the syntactic Merge mechanism for language: A coordinate-based activation likelihood estimation meta-analysis. Neurosci Biobehavioral Reviews 80:646\u0026ndash;656\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZaccarella E, Friederici AD (2015) Syntax in the brain. Brain Mapping: Encyclopedic Ref 461\u0026ndash;468\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu C, Zaccarella E, Friederici AD (2019) Universal neural basis of structure building evidenced by network modulations emerging from Broca\u0026rsquo;s area: The case of Chinese. Hum Brain Mapp 40:1705\u0026ndash;1717\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIndefrey P et al (2001) A neural correlate of syntactic encoding during speech production. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 98, 5933\u0026ndash;5936\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIndefrey P, Hellwig F, Herzog H, Seitz RJ, Hagoort P (2004) Neural responses to the production and comprehension of syntax in identical utterances. Brain Lang 89:312\u0026ndash;319\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMakuuchi M, Bahlmann J, Anwander A, Friederici AD (2009) Segregating the core computational faculty of human language from working memory. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 106, 8362\u0026ndash;8367\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMakuuchi M, Grodzinsky Y, Amunts K, Santi A, Friederici AD (2013) Processing noncanonical sentences in Broca\u0026rsquo;s region: reflections of movement distance and type. Cereb Cortex 23:694\u0026ndash;702\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang P et al (2021) Functional brain plasticity during L1 training on complex sentences: Changes in gamma-band oscillatory activity. Hum Brain Mapp 42:3858\u0026ndash;3870\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen L et al (2023) Frontotemporal effective connectivity revealed a language-general syntactic network for Mandarin Chinese. J Neurolinguistics 66:101127\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTyler LK et al (2010) Preserving Syntactic Processing across the Adult Life Span: The Modulation of the Frontotemporal Language System in the Context of Age-Related Atrophy. Cereb Cortex 20:352\u0026ndash;364\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSkeide MA, Brauer J, Friederici AD (2016) Brain functional and structural predictors of language performance. Cereb Cortex 26:2127\u0026ndash;2139\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBahlmann J, Schubotz RI, Friederici AD (2008) Hierarchical artificial grammar processing engages Broca\u0026rsquo;s area. NeuroImage 42:525\u0026ndash;534\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBahlmann J, Schubotz RI, Mueller JL, Koester D, Friederici AD (2009) Neural circuits of hierarchical visuo-spatial sequence processing. Brain Res 1298:161\u0026ndash;170\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFriederici AD, Bahlmann J, Heim S, Schubotz RI, Anwander A (2006) The brain differentiates human and non-human grammars: Functional localization and structural connectivity. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 103, 2458\u0026ndash;2463\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpitz B, Friederici AD (2003) Interactions of the hippocampal system and the prefrontal cortex in learning language-like rules. NeuroImage 19:1730\u0026ndash;1737\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpitz B, Friederici AD (2004) Brain correlates of language learning: the neuronal dissociation of rule-based versus similarity-based learning. J Neurosci 24:8436\u0026ndash;8440\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpitz B, Friederici AD (2007) Neural basis of processing sequential and hierarchical syntactic structures. Hum Brain Mapp 28:585\u0026ndash;592\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhta S, Fukui N, Sakai KL (2013) Syntactic computation in the human brain: the degree of merger as a key factor. PLoS ONE 8:e56230\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatchin W, Hickok G (2020) The cortical organization of syntax. Cereb Cortex 30:1481\u0026ndash;1498\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatchin W et al (2024) Verbal working memory and syntactic comprehension segregate into the dorsal and ventral streams, respectively. Brain Commun 6:fcae449\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePylkk\u0026auml;nen L (2019) The neural basis of combinatory syntax and semantics. Science 366:62\u0026ndash;66\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLaw R, Pylkk\u0026auml;nen L (2021) Lists with and without syntax: A new approach to measuring the neural processing of syntax. J Neurosci 41:2186\u0026ndash;2196\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatar S, Dirani J, Marantz A, Pylkk\u0026auml;nen L (2021) Left posterior temporal cortex is sensitive to syntax within conceptually matched Arabic expressions. Sci Rep 11:7181\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAcheson DJ, Hagoort P (2013) Stimulating the Brain\u0026rsquo;s Language Network: Syntactic Ambiguity Resolution after TMS to the Inferior Frontal Gyrus and Middle Temporal Gyrus. J Cogn Neurosci 25:1664\u0026ndash;1677\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarreiras M, Pattamadilok C, Meseguer E, Barber H, Devlin JT (2012) Broca\u0026rsquo;s area plays a causal role in morphosyntactic processing. Neuropsychologia 50:816\u0026ndash;820\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoetzee JP et al (2022) Dissociating Language and Thought in Human Reasoning. Brain Sci 13:67\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKroczek LO, Gunter TC, Rysop AU, Friederici AD, Hartwigsen G (2019) Contributions of left frontal and temporal cortex to sentence comprehension: Evidence from simultaneous TMS-EEG. Cortex 115:86\u0026ndash;98\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuhnke P, Meyer L, Friederici AD, Hartwigsen G (2017) Left posterior inferior frontal gyrus is causally involved in reordering during sentence processing. NeuroImage 148:254\u0026ndash;263\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaria-Korina S, THE ROLE OF BROCA\u0026rsquo;S AREA IN SYNTAX: A TMS STUDY ON WRITTEN GREEK LANGUAGE (2015). \u003cem\u003eEuropean Scientific Journal, ESJ\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://eujournal.org/index.php/esj/article/view/6479\u003c/span\u003e\u003cspan address=\"https://eujournal.org/index.php/esj/article/view/6479\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeyer L, Elsner A, Turker S, Kuhnke P, Hartwigsen G (2018) Perturbation of left posterior prefrontal cortex modulates top-down processing in sentence comprehension. NeuroImage 181:598\u0026ndash;604\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSakai KL, Noguchi Y, Takeuchi T, Watanabe E (2002) Selective Priming of Syntactic Processing by Event-Related Transcranial Magnetic Stimulation of Broca\u0026rsquo;s Area. Neuron 35:1177\u0026ndash;1182\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUdd\u0026eacute;n J et al (2008) The inferior frontal cortex in artificial syntax processing: An rTMS study. Brain Res 1224:69\u0026ndash;78\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUdd\u0026eacute;n J, Ingvar M, Hagoort P, Petersson KM (2017) Broca\u0026rsquo;s region: A causal role in implicit processing of grammars with crossed non-adjacent dependencies. Cognition 164:188\u0026ndash;198\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Der Burght CL, Numssen O, Schlaak B, Goucha T, Hartwigsen G (2023) Differential contributions of inferior frontal gyrus subregions to sentence processing guided by intonation. Hum Brain Mapp 44:585\u0026ndash;598\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu J et al (2024) Continuous theta-burst stimulation on the left posterior inferior frontal gyrus perturbs complex syntactic processing stability in mandarin Chinese. Neurobiol Lang 5:608\u0026ndash;627\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao C et al (2025) Continuous theta-burst stimulation demonstrates language-network-specific causal effects on syntactic processing. NeuroImage 306:121014\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMaran M, Friederici AD, Zaccarella E (2022) Syntax through the looking glass: A review on two-word linguistic processing across behavioral, neuroimaging and neurostimulation studies. Neurosci Biobehavioral Reviews 142:104881\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchro\u0026euml;n JAM et al (2023) Causal evidence for a coordinated temporal interplay within the language network. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e 120, e2306279120\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFomenko A, Neudorfer C, Dallapiazza RF, Kalia SK, Lozano AM (2018) Low-intensity ultrasound neuromodulation: An overview of mechanisms and emerging human applications. Brain Stimul 11:1209\u0026ndash;1217\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDarmani G et al (2025) Individualized non-invasive deep brain stimulation of the basal ganglia using transcranial ultrasound stimulation. Nat Commun 16:2693\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO\u0026rsquo;Reilly MA (2024) Exploiting the mechanical effects of ultrasound for noninvasive therapy. Science 385:eadp7206\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSarica C et al (2022) Human Studies of Transcranial Ultrasound neuromodulation: A systematic review of effectiveness and safety. Brain Stimul 15:737\u0026ndash;746\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonti MM, Schnakers C, Korb AS, Bystritsky A, Vespa PM (2016) Non-Invasive Ultrasonic Thalamic Stimulation in Disorders of Consciousness after Severe Brain Injury: A First-in-Man Report. Brain Stimulation: Basic Translational Clin Res Neuromodulation 9:940\u0026ndash;941\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen S-G et al (2020) Transcranial focused ultrasound pulsation suppresses pentylenetetrazol induced epilepsy \u003cem\u003ein vivo\u003c/em\u003e. Brain Stimul 13:35\u0026ndash;46\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYaakub S et al (2024) Non-invasive Ultrasonic Neuromodulation of the Human Nucleus Accumbens Impacts Reward Sensitivit. Preprint at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2024.07.25.605068\u003c/span\u003e\u003cspan address=\"10.1101/2024.07.25.605068\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakajima K et al (2022) A causal role of anterior prefrontal-putamen circuit for response inhibition revealed by transcranial ultrasound stimulation in humans. Cell Rep 40\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiebell P et al (2023) Inhibition of midfrontal theta with transcranial ultrasound explains greater approach versus withdrawal behavior in humans. Brain Stimul 16:1278\u0026ndash;1288\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJung J, Atkinson-Clement C, Kaiser M, Ralph MAL (2025) Transcranial focused ultrasound stimulation of the anterior temporal lobe enhances semantic memory by modulating brain morphology, neurochemistry and neural dynamics. 03.10.642483 Preprint at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2025.03.10.642483\u003c/span\u003e\u003cspan address=\"10.1101/2025.03.10.642483\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThibault S et al (2021) Tool use and language share syntactic processes and neural patterns in the basal ganglia. Science 374:eabe0874\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRatcliff R, McKoon G (2008) The diffusion decision model: theory and data for two-choice decision tasks. Neural Comput 20:873\u0026ndash;922\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWiecki TV, Sofer I, Frank MJ (2013) HDDM: Hierarchical Bayesian estimation of the drift-diffusion model in Python. Front neuroinformatics 7:14\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLegon W et al (2014) Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans. Nat Neurosci 17:322\u0026ndash;329\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUdd\u0026eacute;n J, Bahlmann J (2012) A rostro-caudal gradient of structured sequence processing in the left inferior frontal gyrus. Philosophical Trans Royal Soc B: Biol Sci 367:2023\u0026ndash;2032\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng K et al (2022) Induction of Human Motor Cortex Plasticity by Theta Burst Transcranial Ultrasound Stimulation. Ann Neurol 91:238\u0026ndash;252\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeng K et al (2024) Effects of different sonication parameters of theta burst transcranial ultrasound stimulation on human motor cortex. Brain Stimul 17:258\u0026ndash;268\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJust MA, Carpenter PA, Keller TA, Eddy WF, Thulborn KR (1996) Brain Activation Modulated by Sentence Comprehension. Science 274:114\u0026ndash;116\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMatchin W, Hammerly C, Lau E (2017) The role of the IFG and pSTS in syntactic prediction: Evidence from a parametric study of hierarchical structure in fMRI. Cortex 88:106\u0026ndash;123\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUdd\u0026eacute;n J, M\u0026auml;nnel C (2018) Artificial Grammar Learning and Its Neurobiology in Relation to Language Processing and Development. In: Rueschemeyer S-A, Gaskell MG (eds) The Oxford Handbook of Psycholinguistics. Oxford University Press, pp 755\u0026ndash;783. doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/oxfordhb/9780198786825.013.33\u003c/span\u003e\u003cspan address=\"10.1093/oxfordhb/9780198786825.013.33\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStromswold K, Caplan D, Alpert N, Rauch S (1996) Localization of Syntactic Comprehension by Positron Emission Tomography. Brain Lang 52:452\u0026ndash;473\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Z et al (2025) Cross-species characterization of transcranial ultrasound propagation. Brain Stimul 18:164\u0026ndash;172\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePichardo S, BabelBrain (2023) An Open-Source Application for Prospective Modeling of Transcranial Focused Ultrasound for Neuromodulation Applications. IEEE Trans Ultrason Ferroelectr Freq Control 70:587\u0026ndash;599\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoessler L et al (2009) Automated cortical projection of EEG sensors: anatomical correlation via the international 10\u0026ndash;10 system. NeuroImage 46:64\u0026ndash;72\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScrivener CL, Reader AT (2022) Variability of EEG electrode positions and their underlying brain regions: visualizing gel artifacts from a simultaneous EEG-fMRI dataset. Brain Behav 12:e2476\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaayen RH, Davidson DJ, Bates DM (2008) Mixed-effects modeling with crossed random effects for subjects and items. J Mem Lang 59:390\u0026ndash;412\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaeger TF (2008) Categorical data analysis: Away from ANOVAs (transformation or not) and towards logit mixed models. J Mem Lang 59:434\u0026ndash;446\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarr DJ (2013) Random effects structure for testing interactions in linear mixed-effects models. Front Psychol 4\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBates D, M\u0026auml;chler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models using lme4. J Stat Softw 67:1\u0026ndash;48\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStanislaw H, Todorov N (1999) Calculation of signal detection theory measures. Behav Res Methods Instruments Computers 31:137\u0026ndash;149\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyons IM, Price GR, Vaessen A, Blomert L, Ansari D (2014) Numerical predictors of arithmetic success in grades 1\u0026ndash;6. Dev Sci 17:714\u0026ndash;726\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJeon H-A, Friederici AD (2015) Degree of automaticity and the prefrontal cortex. Trends Cogn Sci 19:244\u0026ndash;250\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1 \u003cem\u003eResults of the Linear Mixed-Effects Model for TUS RT\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"877\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" style=\"width: 209px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"5\" rowspan=\"2\" style=\"width: 409px;\"\u003e\n \u003cp\u003eFixed Effects\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" style=\"width: 259px;\"\u003e\n \u003cp\u003eRandom Effects\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" style=\"width: 129px;\"\u003e\n \u003cp\u003eSubject\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 129px;\"\u003e\n \u003cp\u003eItem\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eb\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eSE\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003et\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 154px;\"\u003e\n \u003cp\u003e95%\u003cem\u003e\u0026nbsp;CI\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cem\u003eVariance\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eICC\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e\u003cem\u003eVariance\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eICC\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eIntercept\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e3.160\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e406.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 154px;\"\u003e\n \u003cp\u003e[3.144, 3.176]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.127\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.070\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eLpTL - LIFG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-0.014\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-2.28\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.028\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 154px;\"\u003e\n \u003cp\u003e[-0.027, -0.002]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003evertex - LpTL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-1.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.236\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 154px;\"\u003e\n \u003cp\u003e[-0.017, 0.004]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eCOM - SP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.049\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-14.82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[-0.055, -0.042]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eLpTL - LIFG : COM - SP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.719\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 154px;\"\u003e\n \u003cp\u003e[-0.007, 0.011]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003evertex - LpTL : COM - SP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-0.005\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-1.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.313\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 154px;\"\u003e\n \u003cp\u003e[-0.014, 0.004]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u003c/em\u003e Model: lmer (\u003cem\u003eRT\u003c/em\u003e ~\u0026nbsp;\u003cem\u003eStimulation region\u003c/em\u003e *\u0026nbsp;\u003cem\u003eStructure type\u003c/em\u003e + (1\u0026nbsp;\u003cem\u003e+ Stimulation region\u0026nbsp;\u003c/em\u003e|subject) + (1 |item)).\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2 \u003cem\u003eResults of the Generalized Linear Mixed-Effects Model for TUS ACC\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"877\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"3\" style=\"width: 209px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"5\" rowspan=\"2\" style=\"width: 412px;\"\u003e\n \u003cp\u003eFixed Effects\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" style=\"width: 256px;\"\u003e\n \u003cp\u003eRandom Effects\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" style=\"width: 128px;\"\u003e\n \u003cp\u003eSubject\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 128px;\"\u003e\n \u003cp\u003eItem\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eb\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eSE\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003ez\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e95%\u003cem\u003e\u0026nbsp;CI\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eVariance\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eICC\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eVariance\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u003cem\u003eICC\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eIntercept\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e2.319\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.124\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e18.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e[2.074, 2.562]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.452\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e1.017\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eLpTL - LIFG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.266\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.09\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e2.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e[0.089, 0.441]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.127\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003evertex - LpTL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.082\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.072\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e1.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.258\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e[-0.059, 0.223]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.015\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eCOM - SP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.625\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.098\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e6.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e[0.434, 0.816]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003eLpTL - LIFG : COM - SP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-0.198\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.119\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e-1.66\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.719\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e[-0.431, 0.035]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 209px;\"\u003e\n \u003cp\u003evertex - LpTL : COM - SP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.025\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.123\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.20\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e0.313\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 156px;\"\u003e\n \u003cp\u003e[-0.216, 0.266]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 64px;\"\u003e\n \u003cp\u003e--\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cem\u003eNote.\u003c/em\u003e Model: glmer (\u003cem\u003eACC\u003c/em\u003e ~\u0026nbsp;\u003cem\u003eStimulation region\u003c/em\u003e *\u0026nbsp;\u003cem\u003eStructure type\u003c/em\u003e + (1\u0026nbsp;\u003cem\u003e+ Stimulation region\u0026nbsp;\u003c/em\u003e|subject) + (1 |item)).\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Sentence processing, Posterior temporal lobe, Hierarchical syntactic processing, Inferior frontal gyrus, Transcranial ultrasound stimulation","lastPublishedDoi":"10.21203/rs.3.rs-8717654/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8717654/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eHierarchical syntactic processing is a fundamental ability for human language processing. Previous studies dispute the left inferior frontal gyrus (LIFG) and the left posterior temporal lobe (LpTL as its possible neural underpinnings, however, direct comparisons of the role of these regions during sentence comprehension are rare. We are among the first to use transcranial ultrasound stimulation (TUS), a non-invasive brain stimulation technique with high precision, to test the causal roles of the different brain regions in processing complex sentences (with embedded relative clauses) and simpler coordinated sentences. Participants completed three lab visits (7 days apart), with LIFG, LpTL, or vertex (control) stimulated each time. Our TUS experiment provides consistent evidence for a causal role of the LIFG in syntactic processing: LIFG TUS significantly decreased d-prime (d\u0026rsquo;) sensitivity, prolonged reaction times (RTs), and impaired evidence accumulation during decision-making revealed by the hierarchical drift-diffusion modeling. Our findings besides demonstrating the TUS efficacy in sentence-level syntactic processing strengthen the crucial role of LIFG (not LpTL) for hierarchical syntactic processing, thereby deepen the understanding of this core language faculty\u0026rsquo;s causal neural mechanisms.\u003c/p\u003e","manuscriptTitle":"Transcranial Ultrasound Stimulation Unravels the Causal Role of the Left Inferior Frontal Gyrus in Syntactic Processing","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-30 14:42:11","doi":"10.21203/rs.3.rs-8717654/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"646d7632-2565-469c-9570-45cfd110caae","owner":[],"postedDate":"January 30th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":61990303,"name":"Biological sciences/Neuroscience/Cognitive neuroscience/Language"},{"id":61990304,"name":"Biological sciences/Psychology/Human behaviour"}],"tags":[],"updatedAt":"2026-02-27T16:16:16+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-30 14:42:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8717654","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8717654","identity":"rs-8717654","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}