Mapping Symptom-General and Symptom-Specific Targets for Transcranial Magnetic Stimulation in Schizophrenia: An Electrical Modeling Meta-Analysis

Mapping Symptom-General and Symptom-Specific Targets for Transcranial Magnetic Stimulation in Schizophrenia: An Electrical Modeling Meta-Analysis | 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 Mapping Symptom-General and Symptom-Specific Targets for Transcranial Magnetic Stimulation in Schizophrenia: An Electrical Modeling Meta-Analysis Indrit Bègue, Lorina Sinanaj, Konstantinos Pallis, Anahita Fazel Dehkordi, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5565115/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Sep, 2025 Read the published version in Molecular Psychiatry → Version 1 posted 11 You are reading this latest preprint version Abstract Negative, positive, and cognitive symptoms of schizophrenia relate to disruptions in partially distinct brain circuits. Although promising, transcranial magnetic stimulation (TMS) strategies across and within symptom domains remain to be established due to TMS protocol heterogeneity. For this, we combined standard meta-analysis with electric field (E-field) modeling to identify stimulation sites where E-field strength associated most significantly with clinical improvement. Standard meta-analysis of randomized, sham-controlled studies in 3,806 patients demonstrated benefit of TMS across symptom domains, regardless of target or protocol. Particularly, TMS significantly improved negative and cognitive symptoms with high-frequency stimulation applied to left prefrontal cortex, whereas positive symptoms improved with low-frequency TMS applied to left temporoparietal cortex. In-depth examination of these results with E-field modeling identified stimulation to left dorsomedial prefrontal cortex (L-DMPFC), left orbitofrontal cortex (L-OFC), and left cerebellar crus II and right lobule IX to be significantly associated with improvement across all symptom domains. Especially, greater overlap of studies’ stimulation sites with L-DMPFC and L-OFC related to improved outcomes. For negative symptoms, E-field distribution in L-DMPFC and L-OFC related most significantly to clinical improvement. Specifically, greater proximity to L-DMPFC stimulation site indicated better outcomes, with at trend significance for L-OFC. In the cognitive domain, E-field distribution in frontopolar cortices and left dorsolateral prefrontal cortex related to clinical improvement. Finally, strongest E-field association with clinical improvement was found in the right cerebellar lobules VIIIA, VIIIB, and IX for positive symptoms. These results support symptom-general and symptom-specific TMS approaches for distinct therapeutic goals towards personalized neuromodulation in schizophrenia. Health sciences/Diseases/Psychiatric disorders/Schizophrenia Biological sciences/Neuroscience Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Schizophrenia is a chronic and debilitating disorder affecting up to 1% of the general population. It presents with diverse clinical manifestations including hallucinations and delusions (positive symptoms), apathy and diminished verbal and emotional expression (negative symptoms), and cognitive deficits in a range of processes including working memory, attention and processing speed. Schizophrenia has the highest individual burden of all psychiatric disorders (1), imposing significant economic costs (2) and reducing life expectancy of up to 15 years (3). Notably, cognitive and negative symptoms are associated with worse psychosocial functioning (4) highlighting the urgent need for more effective treatments. While antipsychotic medications primarily target positive symptoms, they show limited efficacy for negative and cognitive symptoms and may even worsen these deficits, emphasizing the need for alternative approaches like transcranial magnetic stimulation (TMS). TMS modulates neural activity through electromagnetic pulses, and has shown promise in targeting symptom-specific neural circuits (5,6). The heterogeneity of symptom presentation in schizophrenia complicates therapeutic decisions, as current approaches generally lack a framework that addresses specific symptom dimensions. This gap highlights the need to differentiate between symptom-general and symptom-specific targets within TMS therapy. Symptom-general targets could be suited for patients with mixed clinical presentation, while symptom-specific targets could allow for precise interventions focused on particularly distressing or disabling symptoms. For instance, TMS targeting auditory processing circuits could help a patient with persistent hallucinations, whereas stimulation aimed at motivational circuits may benefit someone with severe avolition. However, many patients present with mixed symptom profiles, where both specific and broad symptom domains overlap. Such distinction raises an opportunity to tailor TMS to individual symptom profiles, potentially enabling a more personalized and effective approach. Current TMS studies often rely on standard anatomical or scalp-based targeting, which may fail to account for individual neural circuitry underlying each symptom domain (7,8). Therefore, TMS protocols have been applied to varied sites such as primary auditory cortices (i.e., left superior temporal gyrus; STG) to counter auditory hallucinations (9), to distinct nodes in the cerebellar-prefrontal circuit for negative symptoms (10) and to dorsolateral prefrontal cortex for cognitive symptoms (11) with heterogenous outcomes (12). Evidence from depression studies indicates that aligning TMS sites with neuroimaging-defined targets enhances treatment outcomes, underscoring the potential benefits of individualized TMS targeting (13,14). Nevertheless, a comprehensive examination of TMS stimulation sites across and within all symptom domains in schizophrenia is still lacking and could be essential for identifying optimal brain regions for TMS response for diverse therapeutic goals. Beyond spatial targeting, a second source of the variability in TMS response can result from differences in stimulation parameters, such as coil type, position, and intensity, all of which shape the electric field (E-field) distribution and therapeutic effect. Optimizing these parameters has shown enhanced outcomes in major depressive disorder and auditory verbal hallucinations in schizophrenia (15). Advances in meta-analytic methods using E-field strength from transcranial direct current stimulation (tDCS) have mapped E-field distributions across the brain and linked these to behavioral improvements (e.g., in working memory), identifying neural areas with the highest tDCS-induced effects (16,17). However, no study has yet examined whether TMS variability in schizophrenia could be optimized through integrating E-field modeling with meta-analytic data to identify neural areas linked to improvement in schizophrenia symptom domains. To address these challenges, we capitalized on an established combined method that integrates E-field modeling with a standard meta-analytic approach to identify optimal TMS targets in schizophrenia. Specifically, we first conducted a standard meta-analysis to assess TMS efficacy across and within distinct symptoms of schizophrenia and explored potential factors of heterogeneity such as stimulation site and protocol frequency with subgroup analyses. We then simulated E-field distributions and identified brain regions where TMS-induced E-fields most strongly correlated with symptom improvement in schizophrenia. Finally, we validated our findings by testing whether local proximity to these E-field-identified targets was associated with better clinical outcomes in actual TMS studies. 2. Materials and Methods 2.1. Eligibility criteria and search strategy We aimed to identify optimal non-invasive stimulation sites for treatment across and within the negative, cognitive, and positive symptom domains of schizophrenia. First, we conducted a systematic review following PRISMA protocols (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) (illustrated in Fig. 1 A). We screened the literature with the following eligibility criteria. We included randomized, sham-controlled trials of TMS in adults (≥ 18 years) diagnosed with a schizophrenia spectrum disorder (e.g., schizophrenia, schizoaffective disorder, or psychosis). We only included studies for which the outcome (i.e. change in negative, positive or cognitive domains) was the primary endpoint. In other words, we did not include a study in one domain (e.g., negative) if the study was aimed to study another domain (e.g., positive). Focusing on primary endpoints minimizes confounding factors and follows strict meta-analytic standards. Regarding outcome measurement, for negative symptoms, scales such as the Positive and Negative Syndrome Scale Negative Subscale (PANSS-N) and the Scale for the Assessment of Negative Symptoms (SANS) were employed. Cognitive symptoms were assessed using the Brief Assessment of Cognition in Schizophrenia (BACS) composite score, Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) total score, Neurocognitive Composite Score, Montreal Cognitive Assessment (MoCA), Cognitive Stability Index (CSA), Facial Affect Recognition (PFA) and N-back task accuracy (active) assessment score. Positive symptoms were evaluated using the Schizophrenia Hallucination Rating Scale (SHRS), Hallucination Change Scale (HCS), Positive and Negative Syndrome Scale Positive Subscale (PANSS-P), Auditory Hallucinations Rating Scale (AHRS), and the Scale for the Assessment of Positive Symptoms (SAPS). We excluded studies where TMS was used as add-on concomitant treatment. The systematic review is registered on Prospero ( https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=540178 ). We searched PubMed, EMBASE, PsycINFO, and Web of Science databases using terms related to schizophrenia and TMS. The search was carried out the 21st of September 2022 using the following search string in Pubmed: (“schizophreni*” OR "schizoaffective disorder" OR "schizophreniform disorder" OR "schizophrenia"[MeSH Terms] OR “CHR” OR “Clinical High Risk” OR “Ultra High Risk” OR “UHR” OR "Psychotic Disorders"[MeSH Terms] OR "Psychotic Disorder*") AND ("transcranial magnetic stimulation" OR "TMS" OR "rTMS" OR "theta burst" OR "iTBS" OR “cTBS” OR "transcranial Magnetic Stimulation*"[MeSH Terms]). Search strings used for the other databases are available in Appendix S2. Titles and abstracts of studies found through the aforementioned search strategy were independently reviewed by two authors (KP and IB). Full-text versions of studies considered relevant after the initial screening were then evaluated for eligibility. KP and AFD extracted the following information from each included study: author names, publication year, country, study type, number of participants, average participant age, sex distribution, diagnosis, primary and secondary outcomes, side effects, frequency and intensity of TMS (including the total number of stimuli and treatments), motor threshold percentage (MT%), TMS target, use of neuro-navigation, method of targeting (scalp-based, anatomical, EEG landmarks, Montreal Neurological Institute (MNI) coordinates in mm [x, y, z]), nature of the sham intervention, outcome measure (rating scale), and the post-treatment at first follow-up scores. For studies using scalp-based targeting, MNI coordinates were derived from the OPT-TMS study (18). If any of this data was not reported, the authors were contacted via email by KP to request the missing information. If there was no response, data from graphs were extracted using the GetData Graph Digitizer ( http://getdata-graph-digitizer.com/ ) where available. We reviewed previous meta-analyses for post-treatment outcome data needed to calculate effect sizes and excluded studies where data was ultimately not available. The PRISMA algorithm can be found in Figure S1 . In sum, the search yielded 4,269 articles in total from various databases: PubMed (884 studies), EMBASE (1,039 studies), PsycINFO (864 studies), and Web of Science (1,482 studies). After removing 2,013 duplicate records, 2,256 records were screened. Of these, 2,118 records were excluded based on the title and abstract screening. This left 138 articles to be assessed in full text, of which 46 studies did not meet the eligibility criteria. The reasons for exclusion at this stage were primarily not being randomized controlled trials (RCTs) (31 studies), absence of symptom outcome measures (10 studies), and interventions not involving TMS (5 studies). Of the 92 eligible studies, 17 were excluded due to insufficient data and authors not responding to data requests. Additionally, 15 studies were sourced from the meta-analysis by Wang et al. (19) from non-English reports and 8 studies from the meta-analysis by Lorentzen et al (20). This resulted in 98 studies being included in the final analysis, involving a total of 4,739 subjects. Summary tables of all the studies for each symptom domain can be found in the Supplementary Material. We assessed studies for bias across five domains (articles in non-English languages were not evaluated) using the Cochrane Risk of Bias Tool 2.0 ( https://methods.cochrane.org/risk-bias-2 ) : (A) Randomization process (including allocation sequence generation and concealment), (B) Deviations from intended interventions (bias due to non-protocol interventions), (C) Missing outcome data (dropouts), (D) Outcome measurement (using a validated tool), and (E) Selection of the reported result (alignment with the protocol and methods section). According to the Cochrane Risk of Bias Tool 2.0 guidelines, the highest risk score assigned in any domain determined the overall risk of bias score for each study ( https://methods.cochrane.org/risk-bias-2 ) (Table S4). 2.2. Effect size extraction and standard meta-analysis We evaluated the effect size of TMS by analyzing changes in symptom severity across negative, cognitive, and positive domains of the active arm compared to a sham control (Fig. 1 B). To accurately capture treatment effects, we analyzed the mean changes in symptom severity scores rather than simply the final endpoint score which may be confounded for baseline symptom severity for both control and sham groups. Studies lacking baseline scores were omitted from our analysis (see Appendix S3). For consistency of interpretation, in our meta-analysis a positive mean change across all domains indicates improvement. Thus, for negative and positive symptoms, where a decrease in clinical scores indicates improvement, we calculated the mean change as the initial score (baseline) minus the final score (endpoint). For cognitive symptoms, where an increase in scores is better, we calculated mean change as the final score minus the initial score. To quantify the effect size, we used the standardized mean difference (SMD), calculated as Cohen's d, which normalizes the difference between the treatment and sham mean changes by the pooled standard deviation (SD): $$\:SMD=\frac{Mea{n}_{treatment}-Mea{n}_{sham}}{S{D}_{pooled}}$$ For studies with small sample sizes (total participants < 50), we applied Hedges correction factor to adjust for bias in effect size estimation. We assigned a single SMD to each study to represent the effect size associated with each TMS montage. A positive SMD indicates that TMS has effectively improved symptom severity, while an SMD of zero or negative suggests no improvement. To address the high heterogeneity in TMS studies resulting from variations in participant characteristics, coil specifications, and treatment protocols, we employed a random effects model (21). This model considers both within study and between study variance (tau squared τ²) to calculate the cumulative effect size SMD and 95% confidence intervals (CI). We assessed the significance of this effect size using the corresponding 95% CI and p-value. Heterogeneity was measured using Cochran's Q and the I-squared (I²) statistics, adjusted for τ². We also used Kendall’s rank correlation test to evaluate potential publication bias via funnel plot asymmetry test, by examining standard error versus effect size for each symptom domain. Finally, we performed subgroup analyses focused on location of TMS application and frequency of TMS protocol following previous studies (20). 2.3. Finite Element Modeling (FEM) simulation We simulated E-field distributions to quantify how different TMS parameters, namely coil type, position, orientation and stimulation intensity, affected targeted brain regions for each study (Fig. 1 C). Finite element method (FEM) simulations were conducted using SimNIBS version 4.1. (22) and Python 3.12 to model the E-fields induced by TMS. MNI152 standard brain template was used to generate E-field simulations. This template is derived from 3D brain MRI images of 152 adult brains, including 86 male/66 female brains with an average age = 25.02 ± 4.90 and age range = 18 to 44 years old (23). The MNI152 template has been recently validated for correctly estimating group-level TMS-induced E-fields (24). This template provides a full head coverage including the cerebellum, a region excluded in previous E-field modeling studies. Bilateral TMS studies were simulated by applying the principle of superposition to the E-fields from the left and right hemispheres (25). FEM analyses focused on E-field distributions at targets linked to clinical improvement in the standard meta-analysis. For studies using the EEG 10/20 system, the coordinates were transformed to the standard 10/10 electrode position in SimNIBS. We did not perform FEM simulations for studies using coils or stimulators unsupported in SimNIBS (e.g., Cadwell, YRDCCY-I, Neurosoft Ltd.) as well as those lacking available data on type of coil or method of targeting and inferred coordinates in the original article; see full details in Appendix S4. 2.4. Meta-analytic correlation between E-field distribution and effect size Expanding on established methodology (16), we analyzed the relationship between E-field strength and SMD at each brain location (i.e. gray matter tetrahedral element), quantifying this through the Clinical E-Field Correlation (CEC) (Fig. 1 D). This correlation coefficient serves as an indicator of how TMS-induced E-fields influence clinical changes (i.e. improvement) in schizophrenia. Specifically, positive CEC values suggest that higher E-fields in certain brain regions are associated with symptom improvement. Conversely, negative CEC values indicate that lower E-fields in certain regions correlate with symptom improvement, implying that stronger E-field doses in these areas may not be beneficial. CEC values close to zero suggest no significant relationship between E-field strength and symptom improvement in those regions. For correlation analyses, we used the non-parametric Spearman’s rank correlation test as it is less sensitive to outliers and doesn’t make any assumptions about the data distribution. We calculated CEC values and conducted a two-sided permutation test (p < 0.05) by randomly permuting CEC values across each tetrahedral element 1,000 times to create a null distribution. Comparing actual CEC values to this distribution generated p-values for each element. Heatmaps were then produced to visually represent CEC values, highlighting brain regions consistently linked to improved outcomes across studies. Figures are created with BioRender.com. 2.5. Validation of E-field modeling derived TMS targets We assessed the validity of the E-field modeling-derived TMS targets by examining the overlap between these targets and the actual TMS study sites (13): studies whose actual TMS sites were closer to the optimal target derived from our E-field modeling approach, would be associated with better treatment outcomes than those further away. To this end, we standardized the location of all clinical stimulation sites (i.e. specified using scalp-based measurements, anatomical landmarks, EEG coordinates or Talairach in the original studies) by mapping them to MNI coordinates in mm [x, y, z] for each study. Next, we computed the Euclidean distance between individual actual TMS sites and the optimal targets and correlated this distance with the respective study’s standardized mean difference in clinical outcomes (Fig. 1 E). Associations were tested using the nonparametric Spearman rank-order correlation with a two-sided distribution. For visualization purposes (i.e. to minimize data point overlap from closely situated original sites), jittering was applied by introducing random noise to the observations (26). Correlations were not performed if there were less than 10 studies. We limited the validation analysis to regions adjacent to the primary cortical stimulation site; when effects extended beyond this site (e.g., in the cerebellum), the proximity analysis would not be meaningful given that effects would be driven through indirect (i.e. cortico-cerebellar pathways) stimulation. 3. Results Symptom-general and symptom-specific meta-analyses 3.1. TMS is effective across symptom domains A summary of all standard meta-analysis results can be found in Table 1 . Table 1 Standard meta-analysis results TMS montage n studies; N participants per arm SMD 95% CI z p(z) I² τ² Across symptoms and protocols Overall effect 83 studies 2022 active; 1782 sham 0.39 [0.25, 0.52] 5.51 < 0.00001 75% 0.29 Negative symptoms Main effect 54 studies 1316 active; 1150 sham 0.47 [0.30, 0.65] 5.32 < 0.00001 76% 0.30 subgroup: location L-PFC 36 studies 1011 active; 876 sham 0.60 [0.39, 0.81] 5.54 1 Hz 48 studies 1230 active; 1082 sham 0.52 [0.33, 0.70] 5.41 < 0.00001 77% 0.32 = 1 Hz 6 studies 86 active; 68 sham 0.06 [-0.31, 0.44] 0.32 0.75 21% 0.05 iTBS (L-PFC) 6 studies 117 active; 105 sham 0.95 [0.28, 1.61] 2.77 < 0.001 81% 0.56 10 Hz (L-PFC) 21 studies 599 active; 547 sham 0.52 [0.24, 0.80] 3.67 < 0.001 80% 0.32 20 Hz (L-PFC) 7 studies 195 active; 171 sham 0.56 [0.11, 1.00] 2.44 0.015 75% 0.27 Cognitive deficits Main effect 11 studies 340 active; 306 sham 0.32 [0.16, 0.48] 4.00 < 0.0001 0% 0 subgroup: location L-PFC 9 studies 308 active; 282 sham 0.32 [0.15, 0.49] 3.62 < 0.001 2% 0 Positive symptoms Main effect 18 studies 366 active; 328 sham 0.14 [-0.24, 0.51] 0.72 0.47 82% 0.51 subgroup: location L-TPC 10 studies 160 active; 154 sham 0.23 [0.01, 0.45] 2.02 0.043 0% 0 L-PFC 3 studies 66 active; 66 sham 0.79 [0.44, 1.15] 4.39 1 Hz 7 studies 168 active; 116 sham 0.38 [0.08, 0.69] 2.46 0.01 40% 0.07 = 1 Hz 11 studies 198 active; 162 sham -0.04 [-0.65, 0.57] 0.13 0.89 86% 0.88 = 1 Hz (L-TPC) 9 studies 128 active; 122 sham 0.29 [0.04, 0.54] 2.27 0.023 0 0 Abbreviations: L-PFC = left prefrontal cortex, L-TPC = left temporoparietal cortex, iTBS = intermittent theta burst stimulation The standardized mean difference (SMD) across all symptom domains and protocols was 0.39, indicating a small but significant improvement for patients receiving active TMS compared to sham (see forest plot of effect sizes in Figure S2). Significant heterogeneity was noted, and no evidence of publication bias was observed in the funnel plot asymmetry test (Figure S6A). 3.2. TMS is effective in improving negative symptoms The primary meta-analysis of negative symptoms yielded an SMD of 0.47, indicating a small yet significant improvement in negative symptoms for patients receiving active TMS compared to sham (see forest plot in Figure S3, Table 1 ), with moderate-to-high heterogeneity observed. The funnel plot in Figure S6B showed no evidence of publication bias. In the subgroup analyses, we found that high-frequency (HF) TMS demonstrated a significant effect with an SMD of 0.52, while low-frequency (LF) TMS showed no significant effect (Fig. 2 A). High heterogeneity was observed in the HF subgroup (Table 1 ). Stimulation targeting the left prefrontal cortex (L-PFC) with HF yielded an SMD of 0.60. Within this subgroup, iTBS at the L-PFC achieved the highest effect size (SMD = 0.95), followed by 10 Hz (SMD = 0.52) and 20 Hz (SMD = 0.56). High heterogeneity was observed across these subgroups (Fig. 2 B). TMS applied to the left temporoparietal cortex (L-TPC) and cerebellum did not yield significant effects, though the small number of studies in these subgroups warrants cautious interpretation (Fig. 2 C). 3.3. TMS is effective in improving cognitive symptoms The meta-analysis for cognitive symptoms yielded an SMD of 0.32, indicating a small but significant improvement in cognitive symptoms with active TMS (see forest plot of effect sizes in Figure S4, Table 1 ). Heterogeneity was low, and the funnel plot in Figure S6C showed no evidence of publication bias (Table 1 ). TMS targeting the L-PFC showed a consistent effect with an SMD of 0.32. Low heterogeneity was noted in this subgroup (Fig. 3 ). All studies stimulated at HF, thus a separate subgroup analysis by frequency for this domain was not conducted. 3.4. Target-specific high vs low frequency TMS effects in improving positive symptoms The meta-analysis for positive symptoms did not demonstrate a significant overall effect of TMS (SMD = 0.14; see forest plot in Figure S5, Table 1 ), and high heterogeneity was observed. The funnel plot in Figure S6D indicated no publication bias. However, HF TMS showed a small, significant improvement in positive symptoms with an SMD of 0.38 with moderate heterogeneity, while LF TMS did not produce a significant effect (Fig. 4 A). TMS targeting the L-TPC showed a small but significant effect (SMD = 0.23) with no significant heterogeneity (Table 1 ). LF TMS at this site resulted in a higher effect (SMD = 0.29) with no significant heterogeneity indicating consistent effects across studies (Fig. 4 B). TMS targeting the L-PFC showed a large effect (SMD = 0.79) with no heterogeneity, though the small number of studies in this subgroup requires cautious interpretation (Fig. 4 C). Symptom-general and symptom-specific electric field modeling 3.5. Region-specific E-field requirements for enhanced symptom-general outcomes Building on findings from the standard meta-analysis, finite element modeling was applied to correlate E-field distribution with clinical effect size across symptom domains (n = 56 studies), to identify the optimal site defined as the location associated with the highest clinical improvement. We found significant positive CEC values in the left motor cortex (MNI: [-44, -15, 64], CEC MAX = 0.43, p = 0.001), left dorsomedial prefrontal cortex (L-DMPFC, MNI: [-23, 41, 45], CEC MAX = 0.40, p = 0.002), and the left orbitofrontal cortex (L-OFC, MNI: [-44, 48, -15], CEC MAX = 0.33, p = 0.01), indicating that higher E-field strengths in these areas are associated with improved clinical outcomes (Fig. 5 A). Significant negative CECs were found in the cerebellum at left crus II (MNI: [-33, -73, -47], CEC MIN = -0.32, p = 0.01) and right lobule IX (MNI: [22, -65, -61], CEC MIN = -0.45, p = 0.03), suggesting that lower E-field strengths in these regions could yield better clinical outcomes across symptoms (Fig. 5 B). 3.6. E-field strength in the L-DMPFC and L-OFC linked to improvement in negative symptoms with HF TMS Given the previous meta-analysis results showing that HF TMS targeting the L-PFC improves negative symptoms in schizophrenia (Section 3.2 .), we analyzed the relationship between E-field strength and clinical improvements in studies applying HF TMS to the L-PFC (n = 15 studies) (Fig. 6 A). This analysis revealed significant positive CEC values in the L-DMPFC extending to premotor (MNI: [-20, 36, 41], CEC MAX = 0.83, p = 0.0001) and the L-OFC (MNI: [-30, 37, -15], CEC MAX = 0.83, p = 0.0001). 3.7. E-field strength in the L-DLPFC and bilateral frontopolar cortex linked to improvement in cognitive symptoms with HF TMS Building on our findings that HF TMS targeting the L-PFC significantly reduces cognitive deficits in schizophrenia (Section 3.3 .), we further investigated the association between E-field strength and clinical improvement within this subgroup (n = 8 studies) (Fig. 6 B). The correlation analysis revealed a significant cluster of positive CEC values in the left dorsolateral prefrontal cortex (L-DLPFC, MNI: [-41, 12, 44], CEC MAX = 0.76, p = 0.03). Significant negative CEC values were found in bilateral frontopolar cortex (MNI: [± 8, 47, 12], CEC MIN = -0.74, p = 0.04). 3.8. E-field strength in the cerebellum linked to improvement in positive symptom with LF TMS Given the previous results showing improvement in positive symptoms with LF TMS targeting the L-TPC (Section 3.4 .), we examined the correlation between E-field strength and SMD across these studies (n = 8 studies) for positive symptoms (Fig. 6 C). We found statistically significant CEC clusters in the right cerebellar lobules VIIIA (MNI: [32, -38, -50]), VIIIB (MNI: [16, -50, -56]), and IX (MNI: [15, -49, -49]) with CEC MAX = 0.72, p = 0.04 each. Due to a limited number of studies, the subgroup involving the L-PFC (n = 3 studies) was omitted from E-field modeling, despite showing a significant effect size in the standard meta-analysis. 3.9. Validation of electrical-modeling derived brain targets We hypothesized that studies whose actual targeting sites overlapped to our predicted optimal site derived from our E-field modeling approach would be linked to higher clinical improvement. Across symptoms, proximity to the L-DMPFC and L-OFC targets was significantly associated with larger effect sizes (L-DMPFC: r = -0.37, p = 0.003; L-OFC: r = -0.33, p = 0.009) as illustrated in Fig. 5 A. For negative symptoms, proximity to the optimal L-DMPFC target was strongly linked to better outcomes (r = -0.63, p = 0.01). The L-OFC target demonstrated a similar negative coefficient trending towards significance (r = -0.51, p = 0.05), as shown in Fig. 6 A. Due to limited sample size this analysis was not performed for experiments on cognitive (n = 8 studies) and positive (n = 8 studies) domains. 4. Discussion This study combined meta-analysis and E-field modeling to identify TMS targets for symptom-specific and symptom-general improvements in schizophrenia, pinpointing brain regions associated with symptom relief that support personalized treatment approaches. Across symptom domains, TMS demonstrated a beneficial effect (SMD = 0.39), although responses varied based on target regions and stimulation parameters. For positive symptoms, high-frequency TMS (HF TMS; SMD = 0.38) showed efficacy comparable to the median effect size of antipsychotics (SMD = 0.42) (27). In negative symptoms, targeted HF TMS at the L-PFC produced a medium effect size (SMD = 0.60), similar to clozapine (SMD = 0.62) (27). Notably, intermittent theta burst stimulation (iTBS) at the L-PFC showed the highest TMS effect size observed, with a large effect (SMD = 0.95), however more studies are needed. For cognitive symptoms, TMS showed small but significant cognitive gains in schizophrenia (SMD = 0.32), with an effect size similar to that of cognitive remediation (d = 0.29) (28). These findings place TMS as a complementary approach to traditional antipsychotic and non-pharmacological treatment, with promising effects across symptom domains, particularly for challenging negative and cognitive symptoms. While high-frequency TMS demonstrates comparable efficacy to the median effect size of antipsychotics for positive symptoms, its impact on negative symptoms and cognitive symptoms suggests it could bridge gaps in symptom management. The large effect size observed with iTBS at the L-PFC further underscores TMS's potential to advance personalized, domain-targeted interventions in schizophrenia. Building on these symptom-based effects, our analysis also revealed distinct clinical-electrical correlations, which could guide E-field dosing adjustments for optimal therapeutic outcomes. Specifically, regarding symptom-general management, stronger E-fields in the left motor cortex, L-DMPFC, and L-OFC correlated with greater symptom improvement, suggesting that higher E-field strengths in these regions may enhance outcomes. Proximity to the L-DMPFC and L-OFC also correlated with better results, offering a first validation of their potential as therapeutic targets. In contrast, a negative clinical-electrical correlation in the cerebellum indicated higher effects with lower E-field strengths. This likely reflects the cerebellum’s unique neurofunctional architecture, where an inhibitory cerebellar cortex projects onto excitatory pathways. These findings support a more personalized, circuit-specific approach to TMS, enhancing treatment efficacy in schizophrenia by tailoring E-field dosing to distinct neuroanatomical targets (29). In the negative symptom domain, E-field modeling replicated the L-DMPFC and L-OFC as optimal stimulation sites within the PFC based on maximum E-field strength correlation with clinical improvement (Fig. 6 A). Among these sites, the L-DMPFC exhibited the highest positive clinical-electrical improvement value. Again, proximity of actual study targeting sites to the L-DMPFC, related to improved clinical outcomes, validating this site as a promising therapeutic target for negative symptoms. A similar relationship between actual and optimal sites was also found in the L-OFC, albeit at trending significance. Existing literature has highlighted the role of the L-DMPFC and OFC in the pathology of negative symptoms (30–33) and their potential to reduce negative symptom severity through non-invasive stimulation (34). Studies have shown that stimulation of DMPFC is associated with a significant decrease in negative symptoms, especially affective flattening and anhedonia, in schizophrenia patients after DMPFC-rTMS intervention (34). Recent research has shown that DMPFC-rTMS can alleviate depressive symptoms, particularly anhedonia, by enhancing the connectivity of the reward pathway (35–37). Preliminary studies in schizophrenia have shown promising results in treating negative symptoms and cognitive deficits with TMS targeting the right OFC with low-frequency protocols (38). In the cognitive symptom domain, E-field modeling further pinpointed the L-DLPFC as a critical area where higher E-field strengths were significantly associated with cognitive improvements. This finding supports existing evidence (39) emphasizing the essential role of the L-DLPFC in cognitive enhancement through TMS. Conversely, lower E-field strengths in the bilateral frontopolar cortex were also associated with cognitive gains, suggesting region-specific E-field dosing adjustments may optimize treatment effects. These findings align with recent proposals of individual E-field modeling of TMS-induced effects for personalized dosing (29,40). However, given the limited number of studies addressing cognitive symptoms in schizophrenia, further research with larger and more diverse samples is needed to confirm these findings and to refine TMS targeting strategies aimed at cognitive symptom improvement. For positive symptoms, E-field modeling of TMS targeting the L-TPC revealed significant positive correlations between stronger E-fields in specific cerebellar regions—namely, right lobules VIIIA, VIIIB and IX—and symptom improvement, particularly with inhibitory protocols. The left TPC is connected to the cerebellum via cortico-cerebellar pathways, which support bidirectional communication, essential for prediction coding and sensory feedback control (41,42). Aberrant connectivity patterns in the thalamic-cortico-cerebellar network are linked to positive symptoms in schizophrenia, including delusions and hallucinations (43), which in turn have been linked to disruptions in sensory feedback processing (44–46). These disruptions could be partly explained by the altered cerebellum role in forward modeling, i.e. a neural mechanism where the cerebellum continuously generates and updates predictions about sensory input based on incoming information and past experiences, aiming to minimize the error between expectation and reality (47). In schizophrenia, this predictive mechanism is impaired, leading to misattribution of internally generated thoughts as external stimuli (48,49) even in early stages of psychosis (50–52). Our study has several strengths and limitations. A key strength is the use of E-field modeling to directly link TMS-induced electric fields with clinical outcomes, enhancing our understanding of the neuroanatomical correlates of symptom improvement and informing mechanism-based TMS targeting. This approach also enabled E-field dosing recommendations, suggesting lower E-field strengths for cerebellar regions and higher E-field strengths for cortical areas, thereby supporting the development of personalized dosing protocols tailored to neuroanatomy and symptom profiles. Collectively, these findings challenge the “one-site-fits-all” model and underscore the need for targeted, adaptable, E-field dosed TMS strategies to optimize treatment outcomes. However, there are limitations to consider. First, individual anatomical variations, such as gyrification and cortical thickness, were not captured, although the standardized MNI152 brain template supports group-level reliability (53,54). Future efforts can use the same approach in individualized head models. Significant heterogeneity across studies reflects the natural variability in schizophrenia, enhancing the generalizability of our findings. Further exploration into other sources of heterogeneity, such as differences in study designs, participant characteristics, reported targeting methods and outcome measures, is needed to optimize TMS protocols. Another potential area for future research is the impact of accelerated protocols. The effect sizes observed in this study may be smaller than those reported with multiple daily sessions, which are shown to enhance therapeutic outcomes in depression (55). Lastly, while we focused on core TMS parameters, additional protocol variables were not included (56), as they are not currently supported by E-field modeling tools; however, our approach remains clinically relevant for understanding TMS efficacy across and within schizophrenia symptom domains. In conclusion, despite the abovementioned limitations this work may represent a significant step toward personalized TMS therapy, examining how variations in TMS parameters and symptom-specific circuitry influence treatment outcomes, with implications for individualized TMS interventions in schizophrenia. Declarations Conflict of Interest All authors report no conflicts of interest. Acknowledgment L.S. is supported by an Excellence Scholarship of the Swiss Government (ESKAS). I.B. is supported by the Leenaards Foundation. References Solmi M, Seitidis G, Mavridis D, Correll CU, Dragioti E, Guimond S, et al. Incidence, prevalence, and global burden of schizophrenia - data, with critical appraisal, from the Global Burden of Disease (GBD) 2019. Mol Psychiatry. 2023 Dec;28(12):5319–27. Kadakia A, Catillon M, Fan Q, Williams GR, Marden JR, Anderson A, et al. The Economic Burden of Schizophrenia in the United States. J Clin Psychiatry. 2022 Oct 10;83(6):43278. Hjorthøj C, Stürup AE, McGrath JJ, Nordentoft M. Years of potential life lost and life expectancy in schizophrenia: a systematic review and meta-analysis. Lancet Psychiatry. 2017 Apr 1;4(4):295–301. Kalisova L, Michalec J, Dechterenko F, Silhan P, Hyza M, Chlebovcova M, et al. Impact of cognitive performance and negative symptoms on psychosocial functioning in Czech schizophrenia patients. Schizophrenia. 2023 Jul 17;9(1):1–8. Siddiqi SH, Taylor SF, Cooke D, Pascual-Leone A, George MS, Fox MD. Distinct Symptom-Specific Treatment Targets for Circuit-Based Neuromodulation. Am J Psychiatry. 2020 May;177(5):435–46. Beynel L, Powers JP, Appelbaum LG. Effects of repetitive transcranial magnetic stimulation on resting-state connectivity: A systematic review. NeuroImage. 2020 May 1;211:116596. MacKay MAB, Paylor JW, Wong JTF, Winship IR, Baker GB, Dursun SM. Multidimensional Connectomics and Treatment-Resistant Schizophrenia: Linking Phenotypic Circuits to Targeted Therapeutics. Front Psychiatry [Internet]. 2018 Oct 30 [cited 2024 Nov 6];9. Available from: https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2018.00537/full Bègue I, Kaiser S, Kirschner M. Pathophysiology of negative symptom dimensions of schizophrenia – Current developments and implications for treatment. Neurosci Biobehav Rev. 2020 Sep 1;116:74–88. Marzouk T, Winkelbeiner S, Azizi H, Malhotra AK, Homan P. Transcranial Magnetic Stimulation for Positive Symptoms in Schizophrenia: A Systematic Review. Neuropsychobiology. 2019 Sep 10;79(6):384–96. Brady RO, Gonsalvez I, Lee I, Öngür D, Seidman LJ, Schmahmann JD, et al. Cerebellar-Prefrontal Network Connectivity and Negative Symptoms in Schizophrenia. Am J Psychiatry. 2019 Jul;176(7):512–20. Yi S, Wang Q, Wang W, Hong C, Ren Z. Efficacy of repetitive transcranial magnetic stimulation (rTMS) on negative symptoms and cognitive functioning in schizophrenia: An umbrella review of systematic reviews and meta-analyses. Psychiatry Res. 2024 Mar 1;333:115728. Hyde J, Carr H, Kelley N, Seneviratne R, Reed C, Parlatini V, et al. Efficacy of neurostimulation across mental disorders: systematic review and meta-analysis of 208 randomized controlled trials. Mol Psychiatry. 2022 Jun;27(6):2709–19. Cash RFH, Cocchi L, Lv J, Fitzgerald PB, Zalesky A. Functional Magnetic Resonance Imaging–Guided Personalization of Transcranial Magnetic Stimulation Treatment for Depression. JAMA Psychiatry. 2021 Mar 1;78(3):337–9. Cash RFH, Zalesky A. Personalized and Circuit-Based Transcranial Magnetic Stimulation: Evidence, Controversies, and Opportunities. Biol Psychiatry. 2024 Mar 15;95(6):510–22. Cao Z, Xiao X, Zhao Y, Jiang Y, Xie C, Paillère-Martinot ML, et al. Targeting the pathological network: Feasibility of network-based optimization of transcranial magnetic stimulation coil placement for treatment of psychiatric disorders. Front Neurosci [Internet]. 2023 Jan 4 [cited 2024 Nov 6];16. Available from: https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.1079078/full Wischnewski M, Mantell KE, Opitz A. Identifying regions in prefrontal cortex related to working memory improvement: a novel meta-analytic method using electric field modeling. Neurosci Biobehav Rev. 2021 Aug 18;130:147. Wischnewski M, Berger TA, Opitz A. Meta-modeling the effects of anodal left prefrontal transcranial direct current stimulation on working memory performance. Imaging Neurosci. 2024 Jan 25;2:1–14. Johnson KA, Baig M, Ramsey D, Lisanby SH, Avery D, McDonald WM, et al. Prefrontal rTMS for Treating Depression: Location and Intensity Results from the OPT-TMS Multi-Site Clinical Trial. Brain Stimulat. 2012 Mar 14;6(2):108. Wang J, Zhou Y, Gan H, Pang J, Li H, Wang J, et al. Efficacy Towards Negative Symptoms and Safety of Repetitive Transcranial Magnetic Stimulation Treatment for Patients with Schizophrenia: A Systematic Review. Shanghai Arch Psychiatry. 2017 Apr 25;29(2):61. Lorentzen R, Nguyen TD, McGirr A, Hieronymus F, Østergaard SD. The efficacy of transcranial magnetic stimulation (TMS) for negative symptoms in schizophrenia: a systematic review and meta-analysis. Schizophrenia. 2022 Apr 9;8(1):35. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986 Sep 1;7(3):177–88. Thielscher A, Antunes A, Saturnino GB. Field modeling for transcranial magnetic stimulation: A useful tool to understand the physiological effects of TMS? In: 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) [Internet]. 2015 [cited 2024 Nov 6]. p. 222–5. Available from: https://ieeexplore.ieee.org/document/7318340 Fonov V, Evans AC, Botteron K, Almli CR, McKinstry RC, Collins DL. Unbiased average age-appropriate atlases for pediatric studies. NeuroImage. 2011 Jan 1;54(1):313–27. Cho JY, Van Hoornweder S, Sege CT, Antonucci MU, McTeague LM, Caulfield KA. Template MRI scans reliably approximate individual and group-level tES and TMS electric fields induced in motor and prefrontal circuits. Front Neural Circuits [Internet]. 2023 Sep 6 [cited 2024 Nov 6];17. Available from: https://www.frontiersin.org/journals/neural-circuits/articles/10.3389/fncir.2023.1214959/full Grant IS, Phillips WR. Electromagnetism. John Wiley & Sons; 2013. 555 p. Carr DB, Littlefield RJ, Nicholson WL, Littlefield JS. Scatterplot Matrix Techniques for Large N. J Am Stat Assoc. 1987 Jun 1;82(398):424–36. Huhn M, Nikolakopoulou A, Schneider-Thoma J, Krause M, Samara M, Peter N, et al. Comparative efficacy and tolerability of 32 oral antipsychotics for the acute treatment of adults with multi-episode schizophrenia: a systematic review and network meta-analysis. The Lancet. 2019 Sep;394(10202):939–51. Vita A, Barlati S, Ceraso A, Nibbio G, Ariu C, Deste G, et al. Effectiveness, Core Elements, and Moderators of Response of Cognitive Remediation for Schizophrenia: A Systematic Review and Meta-analysis of Randomized Clinical Trials. JAMA Psychiatry. 2021 Aug 1;78(8):848–58. Numssen O, Kuhnke P, Weise K, Hartwigsen G. Electric-field-based dosing for TMS. Imaging Neurosci. 2024 Mar 11;2:1–12. Lieberman MD, Straccia MA, Meyer ML, Du M, Tan KM. Social, self, (situational), and affective processes in medial prefrontal cortex (MPFC): Causal, multivariate, and reverse inference evidence. Neurosci Biobehav Rev. 2019 Apr 1;99:311–28. Harvey PO, Lepage M. Neural correlates of recognition memory of social information in people with schizophrenia. J Psychiatry Neurosci JPN. 2014 Mar;39(2):97. Ferrari C, Lega C, Vernice M, Tamietto M, Mende-Siedlecki P, Vecchi T, et al. The Dorsomedial Prefrontal Cortex Plays a Causal Role in Integrating Social Impressions from Faces and Verbal Descriptions. Cereb Cortex. 2016 Jan 1;26(1):156–65. Strauss GP, Nuñez A, Ahmed AO, Barchard KA, Granholm E, Kirkpatrick B, et al. The Latent Structure of Negative Symptoms in Schizophrenia. JAMA Psychiatry. 2018 Dec 1;75(12):1271–9. Gan H, Zhu J, Zhuo K, Zhang J, Tang Y, Qian Z, et al. High frequency repetitive transcranial magnetic stimulation of dorsomedial prefrontal cortex for negative symptoms in patients with schizophrenia: A double-blind, randomized controlled trial. Psychiatry Res. 2021 May 1;299:113876. Downar J, Geraci J, Salomons TV, Dunlop K, Wheeler S, McAndrews MP, et al. Anhedonia and Reward-Circuit Connectivity Distinguish Nonresponders from Responders to Dorsomedial Prefrontal Repetitive Transcranial Magnetic Stimulation in Major Depression. Biol Psychiatry. 2014 Aug 1;76(3):176–85. Miron JP, Feffer K, Cash RFH, Derakhshan D, Kim JMS, Fettes P, et al. Safety, tolerability and effectiveness of a novel 20 Hz rTMS protocol targeting dorsomedial prefrontal cortex in major depression: An open-label case series. Brain Stimulat. 2019 Sep 1;12(5):1319–21. Dunlop K, Gaprielian P, Blumberger D, Daskalakis ZJ, Kennedy SH, Giacobbe P, et al. MRI-guided dmPFC-rTMS as a Treatment for Treatment-resistant Major Depressive Disorder. J Vis Exp JoVE. 2015 Aug 11;(102):53129. Hu Q, Jiao X, Zhou J, Tang Y, Zhang T, Song C, et al. Low-frequency repetitive transcranial magnetic stimulation over the right orbitofrontal cortex for patients with first-episode schizophrenia: A randomized, double-blind, sham-controlled trial. Psychiatry Res. 2023 Dec 1;330:115600. Jiang Y, Guo Z, Xing G, He L, Peng H, Du F, et al. Effects of High-Frequency Transcranial Magnetic Stimulation for Cognitive Deficit in Schizophrenia: A Meta-Analysis. Front Psychiatry. 2019 Mar 29;10:135. Caulfield KA, Li X, George MS. A reexamination of motor and prefrontal TMS in tobacco use disorder: Time for personalized dosing based on electric field modeling? Clin Neurophysiol Off J Int Fed Clin Neurophysiol. 2021 Sep;132(9):2199–207. Ramnani N. The primate cortico-cerebellar system: anatomy and function. Nat Rev Neurosci. 2006 Jul;7(7):511–22. Sterzer P, Adams RA, Fletcher P, Frith C, Lawrie SM, Muckli L, et al. The Predictive Coding Account of Psychosis. Biol Psychiatry. 2018 Nov 1;84(9):634–43. Ferri J, Ford JM, Roach BJ, Turner JA, Erp TG van, Voyvodic J, et al. Resting-state thalamic dysconnectivity in schizophrenia and relationships with symptoms. Psychol Med. 2018 Nov;48(15):2492–9. Blakemore SJ, Smith J, Steel R, Johnstone EC, Frith CD. The perception of self-produced sensory stimuli in patients with auditory hallucinations and passivity experiences: evidence for a breakdown in self-monitoring. Psychol Med. 2000 Sep;30(5):1131–9. Synofzik M, Thier P, Leube DT, Schlotterbeck P, Lindner A. Misattributions of agency in schizophrenia are based on imprecise predictions about the sensory consequences of one’s actions. Brain. 2010 Jan 1;133(1):262–71. Horga G, Schatz KC, Abi-Dargham A, Peterson BS. Deficits in Predictive Coding Underlie Hallucinations in Schizophrenia. J Neurosci. 2014 Jun 11;34(24):8072–82. Welniarz Q, Worbe Y, Gallea C. The Forward Model: A Unifying Theory for the Role of the Cerebellum in Motor Control and Sense of Agency. Front Syst Neurosci. 2021 Apr 15;15:644059. Randeniya R, Oestreich LKL, Garrido MI. Sensory prediction errors in the continuum of psychosis. Schizophr Res. 2018 Jan 1;191:109–22. Blakemore SJ, Oakley DA, Frith CD. Delusions of alien control in the normal brain. Neuropsychologia. 2003 Jan 1;41(8):1058–67. Wang H, Guo W, Liu F, Wang G, Lyu H, Wu R, et al. Patients with first-episode, drug-naive schizophrenia and subjects at ultra-high risk of psychosis shared increased cerebellar-default mode network connectivity at rest. Sci Rep. 2016 May 18;6:26124. Bernard JA, Orr JM, Mittal VA. Cerebello-thalamo-cortical networks predict positive symptom progression in individuals at ultra-high risk for psychosis. NeuroImage Clin. 2017 Mar 6;14:622. Cao H, Chén OY, Chung Y, Forsyth JK, McEwen SC, Gee DG, et al. Cerebello-thalamo-cortical hyperconnectivity as a state-independent functional neural signature for psychosis prediction and characterization. Nat Commun. 2018 Sep 21;9:3836. Tzirini M, Chatzikyriakou E, Kouskouras K, Foroglou N, Samaras T, Kimiskidis VK. Electric Field Distribution Induced by TMS: Differences Due to Anatomical Variation. Appl Sci. 2022 Jan;12(9):4509. Opitz A, Windhoff M, Heidemann RM, Turner R, Thielscher A. How the brain tissue shapes the electric field induced by transcranial magnetic stimulation. NeuroImage. 2011 Oct 1;58(3):849–59. Cole EJ, Stimpson KH, Bentzley BS, Gulser M, Cherian K, Tischler C, et al. Stanford Accelerated Intelligent Neuromodulation Therapy for Treatment-Resistant Depression. Am J Psychiatry. 2020 Aug;177(8):716–26. Caulfield KA, Brown JC. The Problem and Potential of TMS’ Infinite Parameter Space: A Targeted Review and Road Map Forward. Front Psychiatry [Internet]. 2022 May 10 [cited 2024 Nov 6];13. Available from: https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2022.867091/full Additional Declarations The authors have declared there is NO conflict of interest to disclose Supplementary Files StudyOverviewCognitiveSymptoms.xlsx Study Overview of Included Randomized Controlled Trials in Cognitive Symptoms StudyOverviewNegativeSymptoms.xlsx Study Overview of Included Randomized Controlled Trials in Negative Symptoms StudyOverviewPositiveSymptoms.xlsx Study Overview of Included Randomized Controlled Trials in Positive Symptoms SupplementaryInformation.docx SUPPLEMENTARY INFORMATION Cite Share Download PDF Status: Published Journal Publication published 22 Sep, 2025 Read the published version in Molecular Psychiatry → Version 1 posted Editorial decision: revise 04 Mar, 2025 Review # 1 received at journal 20 Feb, 2025 Review # 2 received at journal 20 Jan, 2025 Reviewer # 3 agreed at journal 07 Jan, 2025 Reviewer # 2 agreed at journal 06 Jan, 2025 Reviewer # 1 agreed at journal 06 Jan, 2025 Reviewers invited by journal 06 Jan, 2025 Editor assigned by journal 04 Dec, 2024 Submission checks completed at journal 04 Dec, 2024 First submitted to journal 03 Dec, 2024 Unknown event 03 Dec, 2024 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-5565115","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":398421183,"identity":"ed8f2dd5-8cbb-4111-97b5-c03b255f460c","order_by":0,"name":"Indrit Bègue","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7klEQVRIiWNgGAWjYBACNgYGA4YEBhACgg8MDDwgmpmwlgSIFsYZxGhhAGlhgGph5oEK4dXCJ9287cPDHwx5/PxrDB/btt2RkXdvYPxcgM9hMseKZwAdViw5442xcW7bMx7DMweYpWfg0yKRYwzyS+KGG2fMpHPbDvMYzkhgg7sQr5b9N86Y/7YEaZn/gEgtG/h7zJgZgVrkJRgIaUkrZkhIkyiWuMFWLNlz7jCPAU9iszQ+LfIzkjcz/rCxyePvP7zxw4+yw/by7YcPfsanBQokgCgBwjQ4wNhAWAMY8B+A2kushlEwCkbBKBgxAAAgZEUwE5EKdAAAAABJRU5ErkJggg==","orcid":"","institution":"University of Geneva","correspondingAuthor":true,"prefix":"","firstName":"Indrit","middleName":"","lastName":"Bègue","suffix":""},{"id":398421184,"identity":"65296073-371b-4ba1-8504-73b57bdbde66","order_by":1,"name":"Lorina Sinanaj","email":"","orcid":"https://orcid.org/0000-0003-4687-5809","institution":"University of Geneva","correspondingAuthor":false,"prefix":"","firstName":"Lorina","middleName":"","lastName":"Sinanaj","suffix":""},{"id":398421185,"identity":"cc78d743-9219-4c0b-a531-9aba39271ba0","order_by":2,"name":"Konstantinos Pallis","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Konstantinos","middleName":"","lastName":"Pallis","suffix":""},{"id":398421186,"identity":"2353d555-c59a-4f7e-8d63-ab9f0008469a","order_by":3,"name":"Anahita Fazel Dehkordi","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Anahita","middleName":"Fazel","lastName":"Dehkordi","suffix":""},{"id":398421187,"identity":"938e263d-a08b-4a8f-8e9b-a3a2ce84e1d7","order_by":4,"name":"Philippe Huguelet","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Philippe","middleName":"","lastName":"Huguelet","suffix":""},{"id":398421188,"identity":"b35eb56f-df4a-4f47-b888-3508c9fb91c7","order_by":5,"name":"Stefan Kaiser","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Stefan","middleName":"","lastName":"Kaiser","suffix":""}],"badges":[],"createdAt":"2024-12-02 13:45:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5565115/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5565115/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41380-025-03238-z","type":"published","date":"2025-09-22T04:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":73315503,"identity":"57f9155f-b1c7-4fc9-bede-6751d9cf230a","added_by":"auto","created_at":"2025-01-08 19:55:37","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":907706,"visible":true,"origin":"","legend":"\u003cp\u003eRationale for optimal TMS targeting in schizophrenia based on E-field modeling. A. Systematic review of randomized, sham-controlled trials. B. Standard meta-analysis across and within symptom domains, followed by subgroup analysis focused on TMS location and frequency. C. Simulation of E-fields for each study, according to specific TMS parameters given in the study. D. For each brain location (tetrahedral element), the correlation between clinical effect size and E-field strength is measured (CEC). Following the calculation of CEC values, a 2-sided permutation test is employed to statistically assess their significance. All CEC values are displayed on the gray matter volume to identify optimal brain targets where TMS has the most beneficial impact. E. Closer distance between actual and optimal TMS targets is associated with improved clinical effect.\u003c/p\u003e","description":"","filename":"Fig1Pipeline.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/5b265563c0cf4b32fff2505c.jpg"},{"id":73315509,"identity":"71162993-c7f0-477f-88f7-5565e1751c55","added_by":"auto","created_at":"2025-01-08 19:55:38","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2046505,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eForest plots depicting the efficacy of TMS treatment versus sham in negative symptoms of schizophrenia, stratified by frequency (A) and location (B, prefrontal cortex and C, other locations). Abbreviations: PFC = prefrontal cortex, TPC = temporoparietal cortex, iTBS = intermittent theta burst stimulation.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"Fig2NegativeSubgroupForestPlots.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/c506596ed6bf589421c87992.jpg"},{"id":73315511,"identity":"97ead6a3-bc06-49c5-89f1-4320d3bd344b","added_by":"auto","created_at":"2025-01-08 19:55:38","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":241351,"visible":true,"origin":"","legend":"\u003cp\u003eForest plot depicting the efficacy of TMS treatment versus sham in cognitive symptoms of schizophrenia, targeting the left prefrontal cortex. Abbreviations: PFC = prefrontal cortex.\u003c/p\u003e","description":"","filename":"Fig3CognitiveSubgroupForestPlots.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/9b1e61f339221c87fcb670c0.jpg"},{"id":73315507,"identity":"f72efb9a-f6dd-47b8-b440-83e27058f96a","added_by":"auto","created_at":"2025-01-08 19:55:38","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":524855,"visible":true,"origin":"","legend":"\u003cp\u003eForest plot depicting the efficacy of TMS treatment versus sham in positive symptoms of schizophrenia, stratified by frequency (A) and location (B, left temporoparietal cortex, and C, other). Abbreviations: TPC = temporoparietal cortex.\u003c/p\u003e","description":"","filename":"Fig4PositiveSubgroupForestPlots.png","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/b22476fe5d4cfe32d2e33b9c.png"},{"id":73316132,"identity":"f5c5f51d-fa38-485d-bc0a-24f0a2e0eae6","added_by":"auto","created_at":"2025-01-08 20:03:38","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1226743,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of the optimal TMS sites (the locations associated with the highest clinical improvements) across symptom domains, highlighting areas where higher E-field strengths are linked to greater clinical improvement (A), and areas where lower E-field strengths are linked to greater clinical improvement (B). Abbreviations: L-DMPFC = left dorsomedial prefrontal cortex, M1 = Motor cortex, L-OFC = left orbitofrontal cortex, L-Crus = Left Crus, R-Lobule = Right Lobule.\u003c/p\u003e","description":"","filename":"Fig5AcrossSymptomsCEC.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/1a31dd9d337e72797008fc0d.jpg"},{"id":73315520,"identity":"16d632be-cbca-479d-909a-e760d6379f3e","added_by":"auto","created_at":"2025-01-08 19:55:38","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2463199,"visible":true,"origin":"","legend":"\u003cp\u003eIdentification of the optimal TMS sites (the locations associated with the highest clinical improvement) for each symptom domain: negative symptoms (A), cognitive symptoms (B) and positive symptoms (C). Abbreviations: L-DMPFC = left dorsomedial prefrontal cortex, L-OFC = left orbitofrontal cortex, L-DLPFC = left dorsolateral prefrontal cortex, FPC = frontopolar cortex, R-Lobule = right Lobule.\u003c/p\u003e","description":"","filename":"Fig6SymptomSpecificCEC.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/0eb6f96642ded86bfcd82a88.jpg"},{"id":91954653,"identity":"fd9bc50b-ab80-4fd1-8795-a30063eaef27","added_by":"auto","created_at":"2025-09-23 07:08:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8435033,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/489af3f6-9a8a-41b3-ae8e-9d4a17589b41.pdf"},{"id":73315500,"identity":"0adaf5de-cfd0-4877-a20b-07ecf3c39aa3","added_by":"auto","created_at":"2025-01-08 19:55:37","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15826,"visible":true,"origin":"","legend":"Study Overview of Included Randomized Controlled Trials in Cognitive Symptoms","description":"","filename":"StudyOverviewCognitiveSymptoms.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/89a7099829426875677fa6ff.xlsx"},{"id":73316134,"identity":"30ed27d1-47d8-4581-82c1-3fd55a6800b5","added_by":"auto","created_at":"2025-01-08 20:03:38","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":33869,"visible":true,"origin":"","legend":"Study Overview of Included Randomized Controlled Trials in Negative Symptoms","description":"","filename":"StudyOverviewNegativeSymptoms.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/c803870c5c6c3a74abfebbee.xlsx"},{"id":73315513,"identity":"af0d3ef4-f7e5-41d1-8215-8e1cd234ce45","added_by":"auto","created_at":"2025-01-08 19:55:38","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":17830,"visible":true,"origin":"","legend":"Study Overview of Included Randomized Controlled Trials in Positive Symptoms","description":"","filename":"StudyOverviewPositiveSymptoms.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/bd2fc3649d6164a1ab1ba60b.xlsx"},{"id":73315518,"identity":"585a724d-1d92-40db-ad53-2c892de109e8","added_by":"auto","created_at":"2025-01-08 19:55:38","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":5261043,"visible":true,"origin":"","legend":"SUPPLEMENTARY INFORMATION","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-5565115/v1/4863a509b1d6cad89242562c.docx"}],"financialInterests":"The authors have declared there is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose","formattedTitle":"Mapping Symptom-General and Symptom-Specific Targets for Transcranial Magnetic Stimulation in Schizophrenia: An Electrical Modeling Meta-Analysis","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSchizophrenia is a chronic and debilitating disorder affecting up to 1% of the general population. It presents with diverse clinical manifestations including hallucinations and delusions (positive symptoms), apathy and diminished verbal and emotional expression (negative symptoms), and cognitive deficits in a range of processes including working memory, attention and processing speed. Schizophrenia has the highest individual burden of all psychiatric disorders (1), imposing significant economic costs (2) and reducing life expectancy of up to 15 years (3). Notably, cognitive and negative symptoms are associated with worse psychosocial functioning (4) highlighting the urgent need for more effective treatments.\u003c/p\u003e \u003cp\u003eWhile antipsychotic medications primarily target positive symptoms, they show limited efficacy for negative and cognitive symptoms and may even worsen these deficits, emphasizing the need for alternative approaches like transcranial magnetic stimulation (TMS). TMS modulates neural activity through electromagnetic pulses, and has shown promise in targeting symptom-specific neural circuits (5,6). The heterogeneity of symptom presentation in schizophrenia complicates therapeutic decisions, as current approaches generally lack a framework that addresses specific symptom dimensions. This gap highlights the need to differentiate between symptom-general and symptom-specific targets within TMS therapy. Symptom-general targets could be suited for patients with mixed clinical presentation, while symptom-specific targets could allow for precise interventions focused on particularly distressing or disabling symptoms. For instance, TMS targeting auditory processing circuits could help a patient with persistent hallucinations, whereas stimulation aimed at motivational circuits may benefit someone with severe avolition. However, many patients present with mixed symptom profiles, where both specific and broad symptom domains overlap. Such distinction raises an opportunity to tailor TMS to individual symptom profiles, potentially enabling a more personalized and effective approach.\u003c/p\u003e \u003cp\u003eCurrent TMS studies often rely on standard anatomical or scalp-based targeting, which may fail to account for individual neural circuitry underlying each symptom domain (7,8). Therefore, TMS protocols have been applied to varied sites such as primary auditory cortices (i.e., left superior temporal gyrus; STG) to counter auditory hallucinations (9), to distinct nodes in the cerebellar-prefrontal circuit for negative symptoms (10) and to dorsolateral prefrontal cortex for cognitive symptoms (11) with heterogenous outcomes (12). Evidence from depression studies indicates that aligning TMS sites with neuroimaging-defined targets enhances treatment outcomes, underscoring the potential benefits of individualized TMS targeting (13,14). Nevertheless, a comprehensive examination of TMS stimulation sites across and within all symptom domains in schizophrenia is still lacking and could be essential for identifying optimal brain regions for TMS response for diverse therapeutic goals.\u003c/p\u003e \u003cp\u003eBeyond spatial targeting, a second source of the variability in TMS response can result from differences in stimulation parameters, such as coil type, position, and intensity, all of which shape the electric field (E-field) distribution and therapeutic effect. Optimizing these parameters has shown enhanced outcomes in major depressive disorder and auditory verbal hallucinations in schizophrenia (15). Advances in meta-analytic methods using E-field strength from transcranial direct current stimulation (tDCS) have mapped E-field distributions across the brain and linked these to behavioral improvements (e.g., in working memory), identifying neural areas with the highest tDCS-induced effects (16,17). However, no study has yet examined whether TMS variability in schizophrenia could be optimized through integrating E-field modeling with meta-analytic data to identify neural areas linked to improvement in schizophrenia symptom domains.\u003c/p\u003e \u003cp\u003eTo address these challenges, we capitalized on an established combined method that integrates E-field modeling with a standard meta-analytic approach to identify optimal TMS targets in schizophrenia. Specifically, we first conducted a \u003cem\u003estandard\u003c/em\u003e meta-analysis to assess TMS efficacy across and within distinct symptoms of schizophrenia and explored potential factors of heterogeneity such as stimulation site and protocol frequency with subgroup analyses. We then simulated E-field distributions and identified brain regions where TMS-induced E-fields most strongly correlated with symptom improvement in schizophrenia. Finally, we validated our findings by testing whether local proximity to these E-field-identified targets was associated with better clinical outcomes in actual TMS studies.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Eligibility criteria and search strategy\u003c/h2\u003e \u003cp\u003eWe aimed to identify optimal non-invasive stimulation sites for treatment across and within the negative, cognitive, and positive symptom domains of schizophrenia. First, we conducted a systematic review following PRISMA protocols (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) (illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). We screened the literature with the following eligibility criteria. We included randomized, sham-controlled trials of TMS in adults (\u0026ge;\u0026thinsp;18 years) diagnosed with a schizophrenia spectrum disorder (e.g., schizophrenia, schizoaffective disorder, or psychosis). We only included studies for which the outcome (i.e. change in negative, positive or cognitive domains) was the \u003cem\u003eprimary\u003c/em\u003e endpoint. In other words, we did not include a study in one domain (e.g., negative) if the study was aimed to study another domain (e.g., positive). Focusing on primary endpoints minimizes confounding factors and follows strict meta-analytic standards.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eRegarding outcome measurement, for negative symptoms, scales such as the Positive and Negative Syndrome Scale Negative Subscale (PANSS-N) and the Scale for the Assessment of Negative Symptoms (SANS) were employed. Cognitive symptoms were assessed using the Brief Assessment of Cognition in Schizophrenia (BACS) composite score, Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) total score, Neurocognitive Composite Score, Montreal Cognitive Assessment (MoCA), Cognitive Stability Index (CSA), Facial Affect Recognition (PFA) and N-back task accuracy (active) assessment score. Positive symptoms were evaluated using the Schizophrenia Hallucination Rating Scale (SHRS), Hallucination Change Scale (HCS), Positive and Negative Syndrome Scale Positive Subscale (PANSS-P), Auditory Hallucinations Rating Scale (AHRS), and the Scale for the Assessment of Positive Symptoms (SAPS). We excluded studies where TMS was used as add-on concomitant treatment. The systematic review is registered on Prospero (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.crd.york.ac.uk/prospero/display_record.php?RecordID=540178\u003c/span\u003e\u003cspan address=\"https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=540178\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e\u003c/p\u003e \u003cp\u003eWe searched PubMed, EMBASE, PsycINFO, and Web of Science databases using terms related to schizophrenia and TMS. The search was carried out the 21st of September 2022 using the following search string in Pubmed: (\u0026ldquo;schizophreni*\u0026rdquo; OR \"schizoaffective disorder\" OR \"schizophreniform disorder\" OR \"schizophrenia\"[MeSH Terms] OR \u0026ldquo;CHR\u0026rdquo; OR \u0026ldquo;Clinical High Risk\u0026rdquo; OR \u0026ldquo;Ultra High Risk\u0026rdquo; OR \u0026ldquo;UHR\u0026rdquo; OR \"Psychotic Disorders\"[MeSH Terms] OR \"Psychotic Disorder*\") AND (\"transcranial magnetic stimulation\" OR \"TMS\" OR \"rTMS\" OR \"theta burst\" OR \"iTBS\" OR \u0026ldquo;cTBS\u0026rdquo; OR \"transcranial Magnetic Stimulation*\"[MeSH Terms]). Search strings used for the other databases are available in Appendix S2.\u003c/p\u003e \u003cp\u003eTitles and abstracts of studies found through the aforementioned search strategy were independently reviewed by two authors (KP and IB). Full-text versions of studies considered relevant after the initial screening were then evaluated for eligibility. KP and AFD extracted the following information from each included study: author names, publication year, country, study type, number of participants, average participant age, sex distribution, diagnosis, primary and secondary outcomes, side effects, frequency and intensity of TMS (including the total number of stimuli and treatments), motor threshold percentage (MT%), TMS target, use of neuro-navigation, method of targeting (scalp-based, anatomical, EEG landmarks, Montreal Neurological Institute (MNI) coordinates in mm [x, y, z]), nature of the sham intervention, outcome measure (rating scale), and the post-treatment at first follow-up scores. For studies using scalp-based targeting, MNI coordinates were derived from the OPT-TMS study (18). If any of this data was not reported, the authors were contacted via email by KP to request the missing information. If there was no response, data from graphs were extracted using the GetData Graph Digitizer (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://getdata-graph-digitizer.com/\u003c/span\u003e\u003cspan address=\"http://getdata-graph-digitizer.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e where available. We reviewed previous meta-analyses for post-treatment outcome data needed to calculate effect sizes and excluded studies where data was ultimately not available.\u003c/p\u003e \u003cp\u003eThe PRISMA algorithm can be found in Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. In sum, the search yielded 4,269 articles in total from various databases: PubMed (884 studies), EMBASE (1,039 studies), PsycINFO (864 studies), and Web of Science (1,482 studies). After removing 2,013 duplicate records, 2,256 records were screened. Of these, 2,118 records were excluded based on the title and abstract screening. This left 138 articles to be assessed in full text, of which 46 studies did not meet the eligibility criteria. The reasons for exclusion at this stage were primarily not being randomized controlled trials (RCTs) (31 studies), absence of symptom outcome measures (10 studies), and interventions not involving TMS (5 studies). Of the 92 eligible studies, 17 were excluded due to insufficient data and authors not responding to data requests. Additionally, 15 studies were sourced from the meta-analysis by Wang et al. (19) from non-English reports and 8 studies from the meta-analysis by Lorentzen et al (20). This resulted in 98 studies being included in the final analysis, involving a total of 4,739 subjects. Summary tables of all the studies for each symptom domain can be found in the Supplementary Material.\u003c/p\u003e \u003cp\u003eWe assessed studies for bias across five domains (articles in non-English languages were not evaluated) using the Cochrane Risk of Bias Tool 2.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://methods.cochrane.org/risk-bias-2\u003c/span\u003e\u003cspan address=\"https://methods.cochrane.org/risk-bias-2\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e: (A) Randomization process (including allocation sequence generation and concealment), (B) Deviations from intended interventions (bias due to non-protocol interventions), (C) Missing outcome data (dropouts), (D) Outcome measurement (using a validated tool), and (E) Selection of the reported result (alignment with the protocol and methods section). According to the Cochrane Risk of Bias Tool 2.0 guidelines, the highest risk score assigned in any domain determined the overall risk of bias score for each study (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://methods.cochrane.org/risk-bias-2\u003c/span\u003e\u003cspan address=\"https://methods.cochrane.org/risk-bias-2\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e (Table S4).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Effect size extraction and standard meta-analysis\u003c/h2\u003e \u003cp\u003eWe evaluated the effect size of TMS by analyzing changes in symptom severity across negative, cognitive, and positive domains of the active arm compared to a sham control (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). To accurately capture treatment effects, we analyzed the mean changes in symptom severity scores rather than simply the final endpoint score which may be confounded for baseline symptom severity for both control and sham groups. Studies lacking baseline scores were omitted from our analysis (see Appendix S3). For consistency of interpretation, in our meta-analysis a positive mean change across all domains indicates improvement. Thus, for negative and positive symptoms, where a decrease in clinical scores indicates improvement, we calculated the mean change as the initial score (baseline) minus the final score (endpoint). For cognitive symptoms, where an increase in scores is better, we calculated mean change as the final score minus the initial score.\u003c/p\u003e \u003cp\u003eTo quantify the effect size, we used the standardized mean difference (SMD), calculated as Cohen's d, which normalizes the difference between the treatment and sham mean changes by the pooled standard deviation (SD):\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:SMD=\\frac{Mea{n}_{treatment}-Mea{n}_{sham}}{S{D}_{pooled}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eFor studies with small sample sizes (total participants\u0026thinsp;\u0026lt;\u0026thinsp;50), we applied Hedges correction factor to adjust for bias in effect size estimation. We assigned a single SMD to each study to represent the effect size associated with each TMS montage. A positive SMD indicates that TMS has effectively improved symptom severity, while an SMD of zero or negative suggests no improvement.\u003c/p\u003e \u003cp\u003eTo address the high heterogeneity in TMS studies resulting from variations in participant characteristics, coil specifications, and treatment protocols, we employed a random effects model (21). This model considers both within study and between study variance (tau squared τ\u0026sup2;) to calculate the cumulative effect size SMD and 95% confidence intervals (CI). We assessed the significance of this effect size using the corresponding 95% CI and p-value. Heterogeneity was measured using Cochran's Q and the I-squared (I\u0026sup2;) statistics, adjusted for τ\u0026sup2;. We also used Kendall\u0026rsquo;s rank correlation test to evaluate potential publication bias via funnel plot asymmetry test, by examining standard error versus effect size for each symptom domain.\u003c/p\u003e \u003cp\u003eFinally, we performed subgroup analyses focused on location of TMS application and frequency of TMS protocol following previous studies (20).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Finite Element Modeling (FEM) simulation\u003c/h2\u003e \u003cp\u003eWe simulated E-field distributions to quantify how different TMS parameters, namely coil type, position, orientation and stimulation intensity, affected targeted brain regions for each study (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC). Finite element method (FEM) simulations were conducted using SimNIBS version 4.1. (22) and Python 3.12 to model the E-fields induced by TMS. MNI152 standard brain template was used to generate E-field simulations. This template is derived from 3D brain MRI images of 152 adult brains, including 86 male/66 female brains with an average age\u0026thinsp;=\u0026thinsp;25.02\u0026thinsp;\u0026plusmn;\u0026thinsp;4.90 and age range\u0026thinsp;=\u0026thinsp;18 to 44 years old (23). The MNI152 template has been recently validated for correctly estimating group-level TMS-induced E-fields (24). This template provides a full head coverage including the cerebellum, a region excluded in previous E-field modeling studies. Bilateral TMS studies were simulated by applying the principle of superposition to the E-fields from the left and right hemispheres (25).\u003c/p\u003e \u003cp\u003eFEM analyses focused on E-field distributions at targets linked to clinical improvement in the standard meta-analysis. For studies using the EEG 10/20 system, the coordinates were transformed to the standard 10/10 electrode position in SimNIBS. We did not perform FEM simulations for studies using coils or stimulators unsupported in SimNIBS (e.g., Cadwell, YRDCCY-I, Neurosoft Ltd.) as well as those lacking available data on type of coil or method of targeting and inferred coordinates in the original article; see full details in Appendix S4.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Meta-analytic correlation between E-field distribution and effect size\u003c/h2\u003e \u003cp\u003eExpanding on established methodology (16), we analyzed the relationship between E-field strength and SMD at each brain location (i.e. gray matter tetrahedral element), quantifying this through the Clinical E-Field Correlation (CEC) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). This correlation coefficient serves as an indicator of how TMS-induced E-fields influence clinical changes (i.e. improvement) in schizophrenia. Specifically, positive CEC values suggest that higher E-fields in certain brain regions are associated with symptom improvement. Conversely, negative CEC values indicate that lower E-fields in certain regions correlate with symptom improvement, implying that stronger E-field doses in these areas may not be beneficial. CEC values close to zero suggest no significant relationship between E-field strength and symptom improvement in those regions. For correlation analyses, we used the non-parametric Spearman\u0026rsquo;s rank correlation test as it is less sensitive to outliers and doesn\u0026rsquo;t make any assumptions about the data distribution.\u003c/p\u003e \u003cp\u003eWe calculated CEC values and conducted a two-sided permutation test (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) by randomly permuting CEC values across each tetrahedral element 1,000 times to create a null distribution. Comparing actual CEC values to this distribution generated p-values for each element. Heatmaps were then produced to visually represent CEC values, highlighting brain regions consistently linked to improved outcomes across studies. Figures are created with BioRender.com.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Validation of E-field modeling derived TMS targets\u003c/h2\u003e \u003cp\u003eWe assessed the validity of the E-field modeling-derived TMS targets by examining the overlap between these targets and the actual TMS study sites (13): studies whose actual TMS sites were closer to the optimal target derived from our E-field modeling approach, would be associated with better treatment outcomes than those further away. To this end, we standardized the location of all clinical stimulation sites (i.e. specified using scalp-based measurements, anatomical landmarks, EEG coordinates or Talairach in the original studies) by mapping them to MNI coordinates in mm [x, y, z] for each study. Next, we computed the Euclidean distance between individual actual TMS sites and the optimal targets and correlated this distance with the respective study\u0026rsquo;s standardized mean difference in clinical outcomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). Associations were tested using the nonparametric Spearman rank-order correlation with a two-sided distribution. For visualization purposes (i.e. to minimize data point overlap from closely situated original sites), jittering was applied by introducing random noise to the observations (26). Correlations were not performed if there were less than 10 studies. We limited the validation analysis to regions adjacent to the primary cortical stimulation site; when effects extended beyond this site (e.g., in the cerebellum), the proximity analysis would not be meaningful given that effects would be driven through indirect (i.e. cortico-cerebellar pathways) stimulation.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003eSymptom-general and symptom-specific meta-analyses\u003c/p\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.1. TMS is effective across symptom domains\u003c/h2\u003e \u003cp\u003eA summary of all standard meta-analysis results can be found in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStandard meta-analysis results\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTMS montage\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003en\u003c/em\u003e studies;\u003c/p\u003e \u003cp\u003eN participants per arm\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSMD\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ez\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ep(z)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eI\u0026sup2;\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eτ\u0026sup2;\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e \u003cp\u003eAcross symptoms and protocols\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverall effect\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e83 studies\u003c/p\u003e \u003cp\u003e2022 active; 1782 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.25, 0.52]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.00001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e75%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eNegative symptoms\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMain effect\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e54 studies\u003c/p\u003e \u003cp\u003e1316 active; 1150 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.30, 0.65]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.00001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e76%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubgroup: location\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL-PFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36 studies\u003c/p\u003e \u003cp\u003e1011 active; 876 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.39, 0.81]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.00001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e78%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.31\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL-TPC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 studies\u003c/p\u003e \u003cp\u003e96 active; 86 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[-0.27, 0.42]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e20%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCerebellum\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5 studies\u003c/p\u003e \u003cp\u003e109 active; 102 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[-0.64, 0.90]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e81%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.62\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubgroup: frequency\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt; 1 Hz\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48 studies\u003c/p\u003e \u003cp\u003e1230 active; 1082 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.33, 0.70]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.00001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e77%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e= 1 Hz\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 studies\u003c/p\u003e \u003cp\u003e86 active; 68 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[-0.31, 0.44]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e21%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.05\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eiTBS (L-PFC)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 studies\u003c/p\u003e \u003cp\u003e117 active; 105 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.28, 1.61]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e81%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10 Hz (L-PFC)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21 studies\u003c/p\u003e \u003cp\u003e599 active; 547 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.24, 0.80]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e80%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e20 Hz (L-PFC)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 studies\u003c/p\u003e \u003cp\u003e195 active; 171 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.11, 1.00]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e75%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.27\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCognitive deficits\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMain effect\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11 studies\u003c/p\u003e \u003cp\u003e340 active; 306 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.16, 0.48]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubgroup: location\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL-PFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 studies\u003c/p\u003e \u003cp\u003e308 active; 282 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.15, 0.49]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003ePositive symptoms\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMain effect\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e18 studies\u003c/p\u003e \u003cp\u003e366 active; 328 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[-0.24, 0.51]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e82%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.51\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubgroup: location\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL-TPC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10 studies\u003c/p\u003e \u003cp\u003e160 active; 154 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.01, 0.45]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.043\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL-PFC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 studies\u003c/p\u003e \u003cp\u003e66 active; 66 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.44, 1.15]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.00001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003esubgroup: frequency\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt; 1 Hz\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7 studies\u003c/p\u003e \u003cp\u003e168 active; 116 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.08, 0.69]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e40%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e= 1 Hz\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11 studies\u003c/p\u003e \u003cp\u003e198 active; 162 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[-0.65, 0.57]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e86%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.88\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e= 1 Hz (L-TPC)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9 studies\u003c/p\u003e \u003cp\u003e128 active; 122 sham\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e[0.04, 0.54]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.023\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e \u003cp\u003e\u003cem\u003eAbbreviations: L-PFC\u0026thinsp;=\u0026thinsp;left prefrontal cortex, L-TPC\u0026thinsp;=\u0026thinsp;left temporoparietal cortex, iTBS\u0026thinsp;=\u0026thinsp;intermittent theta burst stimulation\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe standardized mean difference (SMD) across all symptom domains and protocols was 0.39, indicating a small but significant improvement for patients receiving active TMS compared to sham (see forest plot of effect sizes in Figure S2). Significant heterogeneity was noted, and no evidence of publication bias was observed in the funnel plot asymmetry test (Figure S6A).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e3.2. TMS is effective in improving negative symptoms\u003c/h2\u003e \u003cp\u003eThe primary meta-analysis of negative symptoms yielded an SMD of 0.47, indicating a small yet significant improvement in negative symptoms for patients receiving active TMS compared to sham (see forest plot in Figure S3, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), with moderate-to-high heterogeneity observed. The funnel plot in Figure S6B showed no evidence of publication bias.\u003c/p\u003e \u003cp\u003eIn the subgroup analyses, we found that high-frequency (HF) TMS demonstrated a significant effect with an SMD of 0.52, while low-frequency (LF) TMS showed no significant effect (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). High heterogeneity was observed in the HF subgroup (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eStimulation targeting the left prefrontal cortex (L-PFC) with HF yielded an SMD of 0.60. Within this subgroup, iTBS at the L-PFC achieved the highest effect size (SMD\u0026thinsp;=\u0026thinsp;0.95), followed by 10 Hz (SMD\u0026thinsp;=\u0026thinsp;0.52) and 20 Hz (SMD\u0026thinsp;=\u0026thinsp;0.56). High heterogeneity was observed across these subgroups (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). TMS applied to the left temporoparietal cortex (L-TPC) and cerebellum did not yield significant effects, though the small number of studies in these subgroups warrants cautious interpretation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.3. TMS is effective in improving cognitive symptoms\u003c/h2\u003e \u003cp\u003eThe meta-analysis for cognitive symptoms yielded an SMD of 0.32, indicating a small but significant improvement in cognitive symptoms with active TMS (see forest plot of effect sizes in Figure S4, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Heterogeneity was low, and the funnel plot in Figure S6C showed no evidence of publication bias (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTMS targeting the L-PFC showed a consistent effect with an SMD of 0.32. Low heterogeneity was noted in this subgroup (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). All studies stimulated at HF, thus a separate subgroup analysis by frequency for this domain was not conducted.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Target-specific high vs low frequency TMS effects in improving positive symptoms\u003c/h2\u003e \u003cp\u003eThe meta-analysis for positive symptoms did not demonstrate a significant overall effect of TMS (SMD\u0026thinsp;=\u0026thinsp;0.14; see forest plot in Figure S5, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), and high heterogeneity was observed. The funnel plot in Figure S6D indicated no publication bias.\u003c/p\u003e \u003cp\u003eHowever, HF TMS showed a small, significant improvement in positive symptoms with an SMD of 0.38 with moderate heterogeneity, while LF TMS did not produce a significant effect (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTMS targeting the L-TPC showed a small but significant effect (SMD\u0026thinsp;=\u0026thinsp;0.23) with no significant heterogeneity (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). LF TMS at this site resulted in a higher effect (SMD\u0026thinsp;=\u0026thinsp;0.29) with no significant heterogeneity indicating consistent effects across studies (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). TMS targeting the L-PFC showed a large effect (SMD\u0026thinsp;=\u0026thinsp;0.79) with no heterogeneity, though the small number of studies in this subgroup requires cautious interpretation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003eSymptom-general and symptom-specific electric field modeling\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.5. Region-specific E-field requirements for enhanced symptom-general outcomes\u003c/h2\u003e \u003cp\u003eBuilding on findings from the standard meta-analysis, finite element modeling was applied to correlate E-field distribution with clinical effect size across symptom domains (n\u0026thinsp;=\u0026thinsp;56 studies), to identify the optimal site defined as the location associated with the highest clinical improvement.\u003c/p\u003e \u003cp\u003eWe found significant positive CEC values in the left motor cortex (MNI: [-44, -15, 64], CEC\u003csub\u003eMAX\u003c/sub\u003e = 0.43, p\u0026thinsp;=\u0026thinsp;0.001), left dorsomedial prefrontal cortex (L-DMPFC, MNI: [-23, 41, 45], CEC\u003csub\u003eMAX\u003c/sub\u003e= 0.40, p\u0026thinsp;=\u0026thinsp;0.002), and the left orbitofrontal cortex (L-OFC, MNI: [-44, 48, -15], CEC\u003csub\u003eMAX\u003c/sub\u003e = 0.33, p\u0026thinsp;=\u0026thinsp;0.01), indicating that higher E-field strengths in these areas are associated with improved clinical outcomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Significant negative CECs were found in the cerebellum at left crus II (MNI: [-33, -73, -47], CEC\u003csub\u003eMIN\u003c/sub\u003e = -0.32, p\u0026thinsp;=\u0026thinsp;0.01) and right lobule IX (MNI: [22, -65, -61], CEC\u003csub\u003eMIN\u003c/sub\u003e = -0.45, p\u0026thinsp;=\u0026thinsp;0.03), suggesting that lower E-field strengths in these regions could yield better clinical outcomes across symptoms (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e3.6. E-field strength in the L-DMPFC and L-OFC linked to improvement in negative symptoms with HF TMS\u003c/p\u003e \u003cp\u003eGiven the previous meta-analysis results showing that HF TMS targeting the L-PFC improves negative symptoms in schizophrenia (Section \u003cspan refid=\"Sec10\" class=\"InternalRef\"\u003e3.2\u003c/span\u003e.), we analyzed the relationship between E-field strength and clinical improvements in studies applying HF TMS to the L-PFC (n\u0026thinsp;=\u0026thinsp;15 studies) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). This analysis revealed significant positive CEC values in the L-DMPFC extending to premotor (MNI: [-20, 36, 41], CEC\u003csub\u003eMAX\u003c/sub\u003e = 0.83, p\u0026thinsp;=\u0026thinsp;0.0001) and the L-OFC (MNI: [-30, 37, -15], CEC\u003csub\u003eMAX\u003c/sub\u003e = 0.83, p\u0026thinsp;=\u0026thinsp;0.0001).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e3.7. E-field strength in the L-DLPFC and bilateral frontopolar cortex linked to improvement in cognitive symptoms with HF TMS\u003c/p\u003e \u003cp\u003eBuilding on our findings that HF TMS targeting the L-PFC significantly reduces cognitive deficits in schizophrenia (Section \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e3.3\u003c/span\u003e.), we further investigated the association between E-field strength and clinical improvement within this subgroup (n\u0026thinsp;=\u0026thinsp;8 studies) (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB). The correlation analysis revealed a significant cluster of positive CEC values in the left dorsolateral prefrontal cortex (L-DLPFC, MNI: [-41, 12, 44], CEC\u003csub\u003eMAX\u003c/sub\u003e = 0.76, p\u0026thinsp;=\u0026thinsp;0.03). Significant negative CEC values were found in bilateral frontopolar cortex (MNI: [\u0026plusmn;\u0026thinsp;8, 47, 12], CEC\u003csub\u003eMIN\u003c/sub\u003e = -0.74, p\u0026thinsp;=\u0026thinsp;0.04).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.8. E-field strength in the cerebellum linked to improvement in positive symptom with LF TMS\u003c/h2\u003e \u003cp\u003eGiven the previous results showing improvement in positive symptoms with LF TMS targeting the L-TPC (Section \u003cspan refid=\"Sec12\" class=\"InternalRef\"\u003e3.4\u003c/span\u003e.), we examined the correlation between E-field strength and SMD across these studies (n\u0026thinsp;=\u0026thinsp;8 studies) for positive symptoms (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC). We found statistically significant CEC clusters in the right cerebellar lobules VIIIA (MNI: [32, -38, -50]), VIIIB (MNI: [16, -50, -56]), and IX (MNI: [15, -49, -49]) with CEC\u003csub\u003eMAX\u003c/sub\u003e = 0.72, p\u0026thinsp;=\u0026thinsp;0.04 each.\u003c/p\u003e \u003cp\u003eDue to a limited number of studies, the subgroup involving the L-PFC (n\u0026thinsp;=\u0026thinsp;3 studies) was omitted from E-field modeling, despite showing a significant effect size in the standard meta-analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.9. Validation of electrical-modeling derived brain targets\u003c/h2\u003e \u003cp\u003eWe hypothesized that studies whose \u003cem\u003eactual\u003c/em\u003e targeting sites overlapped to our \u003cem\u003epredicted\u003c/em\u003e optimal site derived from our E-field modeling approach would be linked to higher clinical improvement.\u003c/p\u003e \u003cp\u003eAcross symptoms, proximity to the L-DMPFC and L-OFC targets was significantly associated with larger effect sizes (L-DMPFC: r = -0.37, p\u0026thinsp;=\u0026thinsp;0.003; L-OFC: r = -0.33, p\u0026thinsp;=\u0026thinsp;0.009) as illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA.\u003c/p\u003e \u003cp\u003eFor negative symptoms, proximity to the optimal L-DMPFC target was strongly linked to better outcomes (r = -0.63, p\u0026thinsp;=\u0026thinsp;0.01). The L-OFC target demonstrated a similar negative coefficient trending towards significance (r = -0.51, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.05), as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA.\u003c/p\u003e \u003cp\u003eDue to limited sample size this analysis was not performed for experiments on cognitive (n\u0026thinsp;=\u0026thinsp;8 studies) and positive (n\u0026thinsp;=\u0026thinsp;8 studies) domains.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis study combined meta-analysis and E-field modeling to identify TMS targets for symptom-specific and symptom-general improvements in schizophrenia, pinpointing brain regions associated with symptom relief that support personalized treatment approaches. Across symptom domains, TMS demonstrated a beneficial effect (SMD\u0026thinsp;=\u0026thinsp;0.39), although responses varied based on target regions and stimulation parameters. For positive symptoms, high-frequency TMS (HF TMS; SMD\u0026thinsp;=\u0026thinsp;0.38) showed efficacy comparable to the median effect size of antipsychotics (SMD\u0026thinsp;=\u0026thinsp;0.42) (27). In negative symptoms, targeted HF TMS at the L-PFC produced a medium effect size (SMD\u0026thinsp;=\u0026thinsp;0.60), similar to clozapine (SMD\u0026thinsp;=\u0026thinsp;0.62) (27). Notably, intermittent theta burst stimulation (iTBS) at the L-PFC showed the highest TMS effect size observed, with a large effect (SMD\u0026thinsp;=\u0026thinsp;0.95), however more studies are needed. For cognitive symptoms, TMS showed small but significant cognitive gains in schizophrenia (SMD\u0026thinsp;=\u0026thinsp;0.32), with an effect size similar to that of cognitive remediation (d\u0026thinsp;=\u0026thinsp;0.29) (28). These findings place TMS as a complementary approach to traditional antipsychotic and non-pharmacological treatment, with promising effects across symptom domains, particularly for challenging negative and cognitive symptoms. While high-frequency TMS demonstrates comparable efficacy to the median effect size of antipsychotics for positive symptoms, its impact on negative symptoms and cognitive symptoms suggests it could bridge gaps in symptom management. The large effect size observed with iTBS at the L-PFC further underscores TMS's potential to advance personalized, domain-targeted interventions in schizophrenia.\u003c/p\u003e \u003cp\u003eBuilding on these symptom-based effects, our analysis also revealed distinct clinical-electrical correlations, which could guide E-field dosing adjustments for optimal therapeutic outcomes. Specifically, regarding symptom-general management, stronger E-fields in the left motor cortex, L-DMPFC, and L-OFC correlated with greater symptom improvement, suggesting that higher E-field strengths in these regions may enhance outcomes. Proximity to the L-DMPFC and L-OFC also correlated with better results, offering a first validation of their potential as therapeutic targets. In contrast, a negative clinical-electrical correlation in the cerebellum indicated higher effects with lower E-field strengths. This likely reflects the cerebellum\u0026rsquo;s unique neurofunctional architecture, where an inhibitory cerebellar cortex projects onto excitatory pathways. These findings support a more personalized, circuit-specific approach to TMS, enhancing treatment efficacy in schizophrenia by tailoring E-field dosing to distinct neuroanatomical targets (29).\u003c/p\u003e \u003cp\u003eIn the negative symptom domain, E-field modeling replicated the L-DMPFC and L-OFC as optimal stimulation sites within the PFC based on maximum E-field strength correlation with clinical improvement (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA). Among these sites, the L-DMPFC exhibited the highest positive clinical-electrical improvement value. Again, proximity of actual study targeting sites to the L-DMPFC, related to improved clinical outcomes, validating this site as a promising therapeutic target for negative symptoms. A similar relationship between actual and optimal sites was also found in the L-OFC, albeit at trending significance. Existing literature has highlighted the role of the L-DMPFC and OFC in the pathology of negative symptoms (30\u0026ndash;33) and their potential to reduce negative symptom severity through non-invasive stimulation (34). Studies have shown that stimulation of DMPFC is associated with a significant decrease in negative symptoms, especially affective flattening and anhedonia, in schizophrenia patients after DMPFC-rTMS intervention (34). Recent research has shown that DMPFC-rTMS can alleviate depressive symptoms, particularly anhedonia, by enhancing the connectivity of the reward pathway (35\u0026ndash;37). Preliminary studies in schizophrenia have shown promising results in treating negative symptoms and cognitive deficits with TMS targeting the right OFC with low-frequency protocols (38).\u003c/p\u003e \u003cp\u003eIn the cognitive symptom domain, E-field modeling further pinpointed the L-DLPFC as a critical area where higher E-field strengths were significantly associated with cognitive improvements. This finding supports existing evidence (39) emphasizing the essential role of the L-DLPFC in cognitive enhancement through TMS. Conversely, lower E-field strengths in the bilateral frontopolar cortex were also associated with cognitive gains, suggesting region-specific E-field dosing adjustments may optimize treatment effects. These findings align with recent proposals of individual E-field modeling of TMS-induced effects for personalized dosing (29,40). However, given the limited number of studies addressing cognitive symptoms in schizophrenia, further research with larger and more diverse samples is needed to confirm these findings and to refine TMS targeting strategies aimed at cognitive symptom improvement.\u003c/p\u003e \u003cp\u003eFor positive symptoms, E-field modeling of TMS targeting the L-TPC revealed significant positive correlations between stronger E-fields in specific cerebellar regions\u0026mdash;namely, right lobules VIIIA, VIIIB and IX\u0026mdash;and symptom improvement, particularly with inhibitory protocols. The left TPC is connected to the cerebellum via cortico-cerebellar pathways, which support bidirectional communication, essential for prediction coding and sensory feedback control (41,42). Aberrant connectivity patterns in the thalamic-cortico-cerebellar network are linked to positive symptoms in schizophrenia, including delusions and hallucinations (43), which in turn have been linked to disruptions in sensory feedback processing (44\u0026ndash;46). These disruptions could be partly explained by the altered cerebellum role in forward modeling, i.e. a neural mechanism where the cerebellum continuously generates and updates predictions about sensory input based on incoming information and past experiences, aiming to minimize the error between expectation and reality (47). In schizophrenia, this predictive mechanism is impaired, leading to misattribution of internally generated thoughts as external stimuli (48,49) even in early stages of psychosis (50\u0026ndash;52).\u003c/p\u003e \u003cp\u003eOur study has several strengths and limitations. A key strength is the use of E-field modeling to directly link TMS-induced electric fields with clinical outcomes, enhancing our understanding of the neuroanatomical correlates of symptom improvement and informing mechanism-based TMS targeting. This approach also enabled E-field dosing recommendations, suggesting lower E-field strengths for cerebellar regions and higher E-field strengths for cortical areas, thereby supporting the development of personalized dosing protocols tailored to neuroanatomy and symptom profiles. Collectively, these findings challenge the \u0026ldquo;one-site-fits-all\u0026rdquo; model and underscore the need for targeted, adaptable, E-field dosed TMS strategies to optimize treatment outcomes.\u003c/p\u003e \u003cp\u003eHowever, there are limitations to consider. First, individual anatomical variations, such as gyrification and cortical thickness, were not captured, although the standardized MNI152 brain template supports group-level reliability (53,54). Future efforts can use the same approach in individualized head models. Significant heterogeneity across studies reflects the natural variability in schizophrenia, enhancing the generalizability of our findings. Further exploration into other sources of heterogeneity, such as differences in study designs, participant characteristics, reported targeting methods and outcome measures, is needed to optimize TMS protocols. Another potential area for future research is the impact of accelerated protocols. The effect sizes observed in this study may be smaller than those reported with multiple daily sessions, which are shown to enhance therapeutic outcomes in depression (55). Lastly, while we focused on core TMS parameters, additional protocol variables were not included (56), as they are not currently supported by E-field modeling tools; however, our approach remains clinically relevant for understanding TMS efficacy across and within schizophrenia symptom domains.\u003c/p\u003e \u003cp\u003eIn conclusion, despite the abovementioned limitations this work may represent a significant step toward personalized TMS therapy, examining how variations in TMS parameters and symptom-specific circuitry influence treatment outcomes, with implications for individualized TMS interventions in schizophrenia.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflict of Interest\u003c/h2\u003e \u003cp\u003eAll authors report no conflicts of interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAcknowledgment\u003c/h2\u003e \u003cp\u003eL.S. is supported by an Excellence Scholarship of the Swiss Government (ESKAS). I.B. is supported by the Leenaards Foundation.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSolmi M, Seitidis G, Mavridis D, Correll CU, Dragioti E, Guimond S, et al. Incidence, prevalence, and global burden of schizophrenia - data, with critical appraisal, from the Global Burden of Disease (GBD) 2019. Mol Psychiatry. 2023 Dec;28(12):5319\u0026ndash;27.\u003c/li\u003e\n\u003cli\u003eKadakia A, Catillon M, Fan Q, Williams GR, Marden JR, Anderson A, et al. The Economic Burden of Schizophrenia in the United States. J Clin Psychiatry. 2022 Oct 10;83(6):43278.\u003c/li\u003e\n\u003cli\u003eHjorth\u0026oslash;j C, St\u0026uuml;rup AE, McGrath JJ, Nordentoft M. Years of potential life lost and life expectancy in schizophrenia: a systematic review and meta-analysis. Lancet Psychiatry. 2017 Apr 1;4(4):295\u0026ndash;301.\u003c/li\u003e\n\u003cli\u003eKalisova L, Michalec J, Dechterenko F, Silhan P, Hyza M, Chlebovcova M, et al. Impact of cognitive performance and negative symptoms on psychosocial functioning in Czech schizophrenia patients. Schizophrenia. 2023 Jul 17;9(1):1\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eSiddiqi SH, Taylor SF, Cooke D, Pascual-Leone A, George MS, Fox MD. Distinct Symptom-Specific Treatment Targets for Circuit-Based Neuromodulation. Am J Psychiatry. 2020 May;177(5):435\u0026ndash;46.\u003c/li\u003e\n\u003cli\u003eBeynel L, Powers JP, Appelbaum LG. Effects of repetitive transcranial magnetic stimulation on resting-state connectivity: A systematic review. NeuroImage. 2020 May 1;211:116596.\u003c/li\u003e\n\u003cli\u003eMacKay MAB, Paylor JW, Wong JTF, Winship IR, Baker GB, Dursun SM. Multidimensional Connectomics and Treatment-Resistant Schizophrenia: Linking Phenotypic Circuits to Targeted Therapeutics. Front Psychiatry [Internet]. 2018 Oct 30 [cited 2024 Nov 6];9. Available from: https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2018.00537/full\u003c/li\u003e\n\u003cli\u003eB\u0026egrave;gue I, Kaiser S, Kirschner M. Pathophysiology of negative symptom dimensions of schizophrenia \u0026ndash; Current developments and implications for treatment. Neurosci Biobehav Rev. 2020 Sep 1;116:74\u0026ndash;88.\u003c/li\u003e\n\u003cli\u003eMarzouk T, Winkelbeiner S, Azizi H, Malhotra AK, Homan P. Transcranial Magnetic Stimulation for Positive Symptoms in Schizophrenia: A Systematic Review. Neuropsychobiology. 2019 Sep 10;79(6):384\u0026ndash;96.\u003c/li\u003e\n\u003cli\u003eBrady RO, Gonsalvez I, Lee I, \u0026Ouml;ng\u0026uuml;r D, Seidman LJ, Schmahmann JD, et al. Cerebellar-Prefrontal Network Connectivity and Negative Symptoms in Schizophrenia. Am J Psychiatry. 2019 Jul;176(7):512\u0026ndash;20.\u003c/li\u003e\n\u003cli\u003eYi S, Wang Q, Wang W, Hong C, Ren Z. Efficacy of repetitive transcranial magnetic stimulation (rTMS) on negative symptoms and cognitive functioning in schizophrenia: An umbrella review of systematic reviews and meta-analyses. Psychiatry Res. 2024 Mar 1;333:115728.\u003c/li\u003e\n\u003cli\u003eHyde J, Carr H, Kelley N, Seneviratne R, Reed C, Parlatini V, et al. Efficacy of neurostimulation across mental disorders: systematic review and meta-analysis of 208 randomized controlled trials. Mol Psychiatry. 2022 Jun;27(6):2709\u0026ndash;19.\u003c/li\u003e\n\u003cli\u003eCash RFH, Cocchi L, Lv J, Fitzgerald PB, Zalesky A. Functional Magnetic Resonance Imaging\u0026ndash;Guided Personalization of Transcranial Magnetic Stimulation Treatment for Depression. JAMA Psychiatry. 2021 Mar 1;78(3):337\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eCash RFH, Zalesky A. Personalized and Circuit-Based Transcranial Magnetic Stimulation: Evidence, Controversies, and Opportunities. Biol Psychiatry. 2024 Mar 15;95(6):510\u0026ndash;22.\u003c/li\u003e\n\u003cli\u003eCao Z, Xiao X, Zhao Y, Jiang Y, Xie C, Paill\u0026egrave;re-Martinot ML, et al. Targeting the pathological network: Feasibility of network-based optimization of transcranial magnetic stimulation coil placement for treatment of psychiatric disorders. Front Neurosci [Internet]. 2023 Jan 4 [cited 2024 Nov 6];16. Available from: https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2022.1079078/full\u003c/li\u003e\n\u003cli\u003eWischnewski M, Mantell KE, Opitz A. Identifying regions in prefrontal cortex related to working memory improvement: a novel meta-analytic method using electric field modeling. Neurosci Biobehav Rev. 2021 Aug 18;130:147.\u003c/li\u003e\n\u003cli\u003eWischnewski M, Berger TA, Opitz A. Meta-modeling the effects of anodal left prefrontal transcranial direct current stimulation on working memory performance. Imaging Neurosci. 2024 Jan 25;2:1\u0026ndash;14.\u003c/li\u003e\n\u003cli\u003eJohnson KA, Baig M, Ramsey D, Lisanby SH, Avery D, McDonald WM, et al. Prefrontal rTMS for Treating Depression: Location and Intensity Results from the OPT-TMS Multi-Site Clinical Trial. Brain Stimulat. 2012 Mar 14;6(2):108.\u003c/li\u003e\n\u003cli\u003eWang J, Zhou Y, Gan H, Pang J, Li H, Wang J, et al. Efficacy Towards Negative Symptoms and Safety of Repetitive Transcranial Magnetic Stimulation Treatment for Patients with Schizophrenia: A Systematic Review. Shanghai Arch Psychiatry. 2017 Apr 25;29(2):61.\u003c/li\u003e\n\u003cli\u003eLorentzen R, Nguyen TD, McGirr A, Hieronymus F, \u0026Oslash;stergaard SD. The efficacy of transcranial magnetic stimulation (TMS) for negative symptoms in schizophrenia: a systematic review and meta-analysis. Schizophrenia. 2022 Apr 9;8(1):35.\u003c/li\u003e\n\u003cli\u003eDerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986 Sep 1;7(3):177\u0026ndash;88.\u003c/li\u003e\n\u003cli\u003eThielscher A, Antunes A, Saturnino GB. Field modeling for transcranial magnetic stimulation: A useful tool to understand the physiological effects of TMS? In: 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) [Internet]. 2015 [cited 2024 Nov 6]. p. 222\u0026ndash;5. Available from: https://ieeexplore.ieee.org/document/7318340\u003c/li\u003e\n\u003cli\u003eFonov V, Evans AC, Botteron K, Almli CR, McKinstry RC, Collins DL. Unbiased average age-appropriate atlases for pediatric studies. NeuroImage. 2011 Jan 1;54(1):313\u0026ndash;27.\u003c/li\u003e\n\u003cli\u003eCho JY, Van Hoornweder S, Sege CT, Antonucci MU, McTeague LM, Caulfield KA. Template MRI scans reliably approximate individual and group-level tES and TMS electric fields induced in motor and prefrontal circuits. Front Neural Circuits [Internet]. 2023 Sep 6 [cited 2024 Nov 6];17. Available from: https://www.frontiersin.org/journals/neural-circuits/articles/10.3389/fncir.2023.1214959/full\u003c/li\u003e\n\u003cli\u003eGrant IS, Phillips WR. Electromagnetism. John Wiley \u0026amp; Sons; 2013. 555 p.\u003c/li\u003e\n\u003cli\u003eCarr DB, Littlefield RJ, Nicholson WL, Littlefield JS. Scatterplot Matrix Techniques for Large N. J Am Stat Assoc. 1987 Jun 1;82(398):424\u0026ndash;36.\u003c/li\u003e\n\u003cli\u003eHuhn M, Nikolakopoulou A, Schneider-Thoma J, Krause M, Samara M, Peter N, et al. Comparative efficacy and tolerability of 32 oral antipsychotics for the acute treatment of adults with multi-episode schizophrenia: a systematic review and network meta-analysis. The Lancet. 2019 Sep;394(10202):939\u0026ndash;51.\u003c/li\u003e\n\u003cli\u003eVita A, Barlati S, Ceraso A, Nibbio G, Ariu C, Deste G, et al. Effectiveness, Core Elements, and Moderators of Response of Cognitive Remediation for Schizophrenia: A Systematic Review and Meta-analysis of Randomized Clinical Trials. JAMA Psychiatry. 2021 Aug 1;78(8):848\u0026ndash;58.\u003c/li\u003e\n\u003cli\u003eNumssen O, Kuhnke P, Weise K, Hartwigsen G. Electric-field-based dosing for TMS. Imaging Neurosci. 2024 Mar 11;2:1\u0026ndash;12.\u003c/li\u003e\n\u003cli\u003eLieberman MD, Straccia MA, Meyer ML, Du M, Tan KM. Social, self, (situational), and affective processes in medial prefrontal cortex (MPFC): Causal, multivariate, and reverse inference evidence. Neurosci Biobehav Rev. 2019 Apr 1;99:311\u0026ndash;28.\u003c/li\u003e\n\u003cli\u003eHarvey PO, Lepage M. Neural correlates of recognition memory of social information in people with schizophrenia. J Psychiatry Neurosci JPN. 2014 Mar;39(2):97.\u003c/li\u003e\n\u003cli\u003eFerrari C, Lega C, Vernice M, Tamietto M, Mende-Siedlecki P, Vecchi T, et al. The Dorsomedial Prefrontal Cortex Plays a Causal Role in Integrating Social Impressions from Faces and Verbal Descriptions. Cereb Cortex. 2016 Jan 1;26(1):156\u0026ndash;65.\u003c/li\u003e\n\u003cli\u003eStrauss GP, Nu\u0026ntilde;ez A, Ahmed AO, Barchard KA, Granholm E, Kirkpatrick B, et al. The Latent Structure of Negative Symptoms in Schizophrenia. JAMA Psychiatry. 2018 Dec 1;75(12):1271\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eGan H, Zhu J, Zhuo K, Zhang J, Tang Y, Qian Z, et al. High frequency repetitive transcranial magnetic stimulation of dorsomedial prefrontal cortex for negative symptoms in patients with schizophrenia: A double-blind, randomized controlled trial. Psychiatry Res. 2021 May 1;299:113876.\u003c/li\u003e\n\u003cli\u003eDownar J, Geraci J, Salomons TV, Dunlop K, Wheeler S, McAndrews MP, et al. Anhedonia and Reward-Circuit Connectivity Distinguish Nonresponders from Responders to Dorsomedial Prefrontal Repetitive Transcranial Magnetic Stimulation in Major Depression. Biol Psychiatry. 2014 Aug 1;76(3):176\u0026ndash;85.\u003c/li\u003e\n\u003cli\u003eMiron JP, Feffer K, Cash RFH, Derakhshan D, Kim JMS, Fettes P, et al. Safety, tolerability and effectiveness of a novel 20 Hz rTMS protocol targeting dorsomedial prefrontal cortex in major depression: An open-label case series. Brain Stimulat. 2019 Sep 1;12(5):1319\u0026ndash;21.\u003c/li\u003e\n\u003cli\u003eDunlop K, Gaprielian P, Blumberger D, Daskalakis ZJ, Kennedy SH, Giacobbe P, et al. MRI-guided dmPFC-rTMS as a Treatment for Treatment-resistant Major Depressive Disorder. J Vis Exp JoVE. 2015 Aug 11;(102):53129.\u003c/li\u003e\n\u003cli\u003eHu Q, Jiao X, Zhou J, Tang Y, Zhang T, Song C, et al. Low-frequency repetitive transcranial magnetic stimulation over the right orbitofrontal cortex for patients with first-episode schizophrenia: A randomized, double-blind, sham-controlled trial. Psychiatry Res. 2023 Dec 1;330:115600.\u003c/li\u003e\n\u003cli\u003eJiang Y, Guo Z, Xing G, He L, Peng H, Du F, et al. Effects of High-Frequency Transcranial Magnetic Stimulation for Cognitive Deficit in Schizophrenia: A Meta-Analysis. Front Psychiatry. 2019 Mar 29;10:135.\u003c/li\u003e\n\u003cli\u003eCaulfield KA, Li X, George MS. A reexamination of motor and prefrontal TMS in tobacco use disorder: Time for personalized dosing based on electric field modeling? Clin Neurophysiol Off J Int Fed Clin Neurophysiol. 2021 Sep;132(9):2199\u0026ndash;207.\u003c/li\u003e\n\u003cli\u003eRamnani N. The primate cortico-cerebellar system: anatomy and function. Nat Rev Neurosci. 2006 Jul;7(7):511\u0026ndash;22.\u003c/li\u003e\n\u003cli\u003eSterzer P, Adams RA, Fletcher P, Frith C, Lawrie SM, Muckli L, et al. The Predictive Coding Account of Psychosis. Biol Psychiatry. 2018 Nov 1;84(9):634\u0026ndash;43.\u003c/li\u003e\n\u003cli\u003eFerri J, Ford JM, Roach BJ, Turner JA, Erp TG van, Voyvodic J, et al. Resting-state thalamic dysconnectivity in schizophrenia and relationships with symptoms. Psychol Med. 2018 Nov;48(15):2492\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eBlakemore SJ, Smith J, Steel R, Johnstone EC, Frith CD. The perception of self-produced sensory stimuli in patients with auditory hallucinations and passivity experiences: evidence for a breakdown in self-monitoring. Psychol Med. 2000 Sep;30(5):1131\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eSynofzik M, Thier P, Leube DT, Schlotterbeck P, Lindner A. Misattributions of agency in schizophrenia are based on imprecise predictions about the sensory consequences of one\u0026rsquo;s actions. Brain. 2010 Jan 1;133(1):262\u0026ndash;71.\u003c/li\u003e\n\u003cli\u003eHorga G, Schatz KC, Abi-Dargham A, Peterson BS. Deficits in Predictive Coding Underlie Hallucinations in Schizophrenia. J Neurosci. 2014 Jun 11;34(24):8072\u0026ndash;82.\u003c/li\u003e\n\u003cli\u003eWelniarz Q, Worbe Y, Gallea C. The Forward Model: A Unifying Theory for the Role of the Cerebellum in Motor Control and Sense of Agency. Front Syst Neurosci. 2021 Apr 15;15:644059.\u003c/li\u003e\n\u003cli\u003eRandeniya R, Oestreich LKL, Garrido MI. Sensory prediction errors in the continuum of psychosis. Schizophr Res. 2018 Jan 1;191:109\u0026ndash;22.\u003c/li\u003e\n\u003cli\u003eBlakemore SJ, Oakley DA, Frith CD. Delusions of alien control in the normal brain. Neuropsychologia. 2003 Jan 1;41(8):1058\u0026ndash;67.\u003c/li\u003e\n\u003cli\u003eWang H, Guo W, Liu F, Wang G, Lyu H, Wu R, et al. Patients with first-episode, drug-naive schizophrenia and subjects at ultra-high risk of psychosis shared increased cerebellar-default mode network connectivity at rest. Sci Rep. 2016 May 18;6:26124.\u003c/li\u003e\n\u003cli\u003eBernard JA, Orr JM, Mittal VA. Cerebello-thalamo-cortical networks predict positive symptom progression in individuals at ultra-high risk for psychosis. NeuroImage Clin. 2017 Mar 6;14:622.\u003c/li\u003e\n\u003cli\u003eCao H, Ch\u0026eacute;n OY, Chung Y, Forsyth JK, McEwen SC, Gee DG, et al. Cerebello-thalamo-cortical hyperconnectivity as a state-independent functional neural signature for psychosis prediction and characterization. Nat Commun. 2018 Sep 21;9:3836.\u003c/li\u003e\n\u003cli\u003eTzirini M, Chatzikyriakou E, Kouskouras K, Foroglou N, Samaras T, Kimiskidis VK. Electric Field Distribution Induced by TMS: Differences Due to Anatomical Variation. Appl Sci. 2022 Jan;12(9):4509.\u003c/li\u003e\n\u003cli\u003eOpitz A, Windhoff M, Heidemann RM, Turner R, Thielscher A. How the brain tissue shapes the electric field induced by transcranial magnetic stimulation. NeuroImage. 2011 Oct 1;58(3):849\u0026ndash;59.\u003c/li\u003e\n\u003cli\u003eCole EJ, Stimpson KH, Bentzley BS, Gulser M, Cherian K, Tischler C, et al. Stanford Accelerated Intelligent Neuromodulation Therapy for Treatment-Resistant Depression. Am J Psychiatry. 2020 Aug;177(8):716\u0026ndash;26.\u003c/li\u003e\n\u003cli\u003eCaulfield KA, Brown JC. The Problem and Potential of TMS\u0026rsquo; Infinite Parameter Space: A Targeted Review and Road Map Forward. Front Psychiatry [Internet]. 2022 May 10 [cited 2024 Nov 6];13. Available from: https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2022.867091/full\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"molecular-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"mp","sideBox":"Learn more about [Molecular Psychiatry](http://www.nature.com/mp/)","snPcode":"41380","submissionUrl":"https://mts-mp.nature.com/cgi-bin/main.plex","title":"Molecular Psychiatry","twitterHandle":"@molpsychiatry","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5565115/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5565115/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eNegative, positive, and cognitive symptoms of schizophrenia relate to disruptions in partially distinct brain circuits. Although promising, transcranial magnetic stimulation (TMS) strategies across and within symptom domains remain to be established due to TMS protocol heterogeneity. For this, we combined standard meta-analysis with electric field (E-field) modeling to identify stimulation sites where E-field strength associated most significantly with clinical improvement. Standard meta-analysis of randomized, sham-controlled studies in 3,806 patients demonstrated benefit of TMS across symptom domains, regardless of target or protocol. Particularly, TMS significantly improved negative and cognitive symptoms with high-frequency stimulation applied to left prefrontal cortex, whereas positive symptoms improved with low-frequency TMS applied to left temporoparietal cortex. In-depth examination of these results with E-field modeling identified stimulation to left dorsomedial prefrontal cortex (L-DMPFC), left orbitofrontal cortex (L-OFC), and left cerebellar crus II and right lobule IX to be significantly associated with improvement across all symptom domains. Especially, greater overlap of studies\u0026rsquo; stimulation sites with L-DMPFC and L-OFC related to improved outcomes. For negative symptoms, E-field distribution in L-DMPFC and L-OFC related most significantly to clinical improvement. Specifically, greater proximity to L-DMPFC stimulation site indicated better outcomes, with at trend significance for L-OFC. In the cognitive domain, E-field distribution in frontopolar cortices and left dorsolateral prefrontal cortex related to clinical improvement. Finally, strongest E-field association with clinical improvement was found in the right cerebellar lobules VIIIA, VIIIB, and IX for positive symptoms. These results support symptom-general and symptom-specific TMS approaches for distinct therapeutic goals towards personalized neuromodulation in schizophrenia.\u003c/p\u003e","manuscriptTitle":"Mapping Symptom-General and Symptom-Specific Targets for Transcranial Magnetic Stimulation in Schizophrenia: An Electrical Modeling Meta-Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-08 19:55:32","doi":"10.21203/rs.3.rs-5565115/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2025-03-04T10:11:59+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-02-20T08:12:44+00:00","index":1,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-01-20T13:06:28+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-01-07T11:19:25+00:00","index":3,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-01-06T20:45:09+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-01-06T18:39:26+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2025-01-06T16:45:32+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-12-04T13:08:01+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-12-04T11:38:48+00:00","index":"","fulltext":""},{"type":"submitted","content":"Molecular Psychiatry","date":"2024-12-03T15:35:52+00:00","index":"","fulltext":""},{"type":"checksFailed","content":"","date":"2024-12-03T12:02:34+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"molecular-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"mp","sideBox":"Learn more about [Molecular Psychiatry](http://www.nature.com/mp/)","snPcode":"41380","submissionUrl":"https://mts-mp.nature.com/cgi-bin/main.plex","title":"Molecular Psychiatry","twitterHandle":"@molpsychiatry","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"4d0283e5-fc88-426c-9ce7-ac0da6a11d01","owner":[],"postedDate":"January 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":42444468,"name":"Health sciences/Diseases/Psychiatric disorders/Schizophrenia"},{"id":42444469,"name":"Biological sciences/Neuroscience"}],"tags":[],"updatedAt":"2025-09-23T07:08:10+00:00","versionOfRecord":{"articleIdentity":"rs-5565115","link":"https://doi.org/10.1038/s41380-025-03238-z","journal":{"identity":"molecular-psychiatry","isVorOnly":false,"title":"Molecular Psychiatry"},"publishedOn":"2025-09-22 04:00:00","publishedOnDateReadable":"September 22nd, 2025"},"versionCreatedAt":"2025-01-08 19:55:32","video":"","vorDoi":"10.1038/s41380-025-03238-z","vorDoiUrl":"https://doi.org/10.1038/s41380-025-03238-z","workflowStages":[]},"version":"v1","identity":"rs-5565115","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5565115","identity":"rs-5565115","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}