Spatially constrained ICA from a language task predicts Sensorimotor Network in patients with tumors

Spatially constrained ICA from a language task predicts Sensorimotor Network in patients with tumors | 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 Research Article Spatially constrained ICA from a language task predicts Sensorimotor Network in patients with tumors Eva Calderón-Rubio, Víctor Costumero, Vicente Belloch, César Ávila This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8287975/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Background Although resting-state fMRI is a promising alternative to task-fMRI in presurgical mapping, protocols to localize simultaneously and accurately both language and motor areas are still missing. Purpose To propose a methodology based on the connectivity analysis to identify motor and language areas using a single 6-minute fMRI language task in presurgical patients with space-occupying lesions. Methods In a retrospective study and using a language fMRI task (verb generation), we established limits of the motor cortex in 40 presurgical patients. Single-subject spatially-constrained ICA was performed on verb generation scans to extract somatomotor network. Sensitivity, and specificity between predicted SMN and hand motor task were calculated. Variables effect was analyzed through ANOVA. Results Forty patients (mean age, 40.50 ± 13.99 [standard deviation]; 21 men) diagnosed with a space-occupying lesion were included. Somatomotor network extracted from spatially-constrained ICA and language lateralization from task-fMRI were identified in 40/40 (100%) patients. Using the motor task as reference standard, ipsilesional voxel-to-voxel mean sensitivity, and specificity for spatially-constrainted ICA at voxel level was 78%, and 75%, respectively. No significant differences were found between ipsilesional and contralesional hemispheres in sensitivity or specificity. Right contralesional hemisphere, higher scanner field strength, and typical language lateralization, showed higher sensitivity values. Conclusions Our study demonstrates the possibility of identifying somatomotor network connectivity from a language task while determining language lateralization in patients with space-occupying lesions. Our technique is a promising alternative to reduce scanning time and cost as only one fMRI sequence is needed to determine the two main eloquent functions. resting-state fMRI motor cortex ICA tumor Figures Figure 1 Figure 2 Figure 3 1. Introduction Surgical resection of brain lesions requires balancing maximal tumor removal with the preservation of cognitive and sensorimotor functions, aiming to avoid permanent postoperative deficits (Lakhani et al., 2023 ). fMRI is a widely used non-invasive tool for presurgical planning, particularly for mapping language and motor functions (Buchbinder, 2016 ). However, task-based fMRI (TB) presents limitations: results depend heavily on patient performance, posing challenges for individuals with cognitive impairment or claustrophobia, and the procedure is costly, time-consuming, and demands active patient cooperation (Leuthardt et al., 2015; Pujol et al., 1998 ). To address the limitations of TB, resting-state (RS) paradigms have emerged as a promising alternative (Parker Jones et al., 2017 ; Sparacia et al., 2021 ). RS offers several advantages: it can reveal multiple functional networks beyond language and motor systems (Seitzman et al., 2019 ), and it does not require active patient participation (Dierker et al., 2017 ). While RS has shown encouraging results for motor mapping (Qiu et al., 2014 ), evidence for language mapping remains limited and less consistent (Sair et al., 2017 ). RS often shows weaker language lateralization and is less sensitive in cases with atypical lateralization, which may lead to misleading presurgical decisions (Branco et al., 2016 ; J. Lu et al., 2017 ; Ott et al., 2021 ; Rolinski et al., 2020 ; Sparacia et al., 2019 ; Tie et al., 2014 ). Given the difficulty of using RS to determine language lateralization, an alternative is to use a single language task and apply ICA to identify SMN (Beheshtian et al., 2021 ; Jurkiewicz et al., 2018 ). However, limitations remain: Jurkiewicz et al. ( 2018 ) performs a seed-based analysis of normalized data in a healthy subject sample that can be limiting when working in native space of presurgical patients diagnosed with space-occupying lesions; and Beheshtian et al. ( 2021 ) did not detect SMN independent components (IC) in all participants and reported RS–TB correspondence in only 75% of cases. Our objective in this study is to overcome the above-mentioned limitations, and to do so we propose a novel methodology that uses a language task-based paradigm to pursue three goals: (1) identify language lateralization; (2) predict SMN activation from a language task; and (3) assess the sensitivity and specificity of these results in comparison with a motor fMRI task (Qiu et al., 2017 ; Qiu et al., 2014 ; Rosazza et al., 2014 ). To address the shortcomings of previous approaches, we apply spatially constrained ICA (scICA) to language-task preprocessed scans. scICA automatically extracts the ICs of interest using a predefined spatial template that serves as a constraint on the source matrix (Lin et al., 2010 ; Lu & Rajapakse, 2005 ). Our main hypothesis is that a language task can be used not only for presurgical language mapping but also to accurately map motor areas of the brain using ICA. 2. Materials and Methods 2.1. Participants. A group of 40 patients (21 males, ages 17–78 years, 40.50 ± 13.99) was selected retrospectively (2009–2022) from a pool of 288 cases for meeting the following inclusion criteria: (1) a space-occupying brain lesion candidate for neurosurgery; (2) a motor task for hand movement (HMT); and (3) a verb-generation task (VGT). Exclusion criteria were incomplete tasks, motion artifacts, or poor in-scanner performance. Demographic data, pathology information, and other variables of interest are reported in Table 1 . Prior to scanning, all patients signed an informed consent form, providing written consent to participate in the study. The study protocol was approved by the university Ethics Committee. Table 1 Demographic data, tumor location, pathologic anatomy analysis results, language lateralization (LI), and type of scanner. Parameters Value Midline shift 13 Ipsilesional central sulcus displacement 8 Lesion location Hemisphere: left / right 29 / 11 Frontal lesion 30 Parietal lesion 2 Temporoparietal lesion 2 Frontotemporal lesion 2 Frontoparietal lesion 4 Anatomic pathology result Oligodendroglioma 5 Glioma 5 Multiform glioblastoma 3 Anaplastic astrocytoma 2 Diffuse astrocytoma 4 Oligoastrocytoma 1 Metastatic lesion 2 Meningioma 2 Cavernoma 2 Cavernous hemangioma 1 Hemangiopericytoma 1 DNET 2 Undetermined Space-occupying brain lesion 10 Scanner Siemens TrioTim 3T 14 Siemens Sonata 1.5T 14 Philips Achieva 3T 12 Note: values are number of patients. 2.2. Image Acquisition Three different scanners were used: Siemens Magnetom Trio 3T, Siemens Magnetom Sonata 1.5T and Philips Achieva 3T X-series. First, a 3D structural MRI was acquired using a T1-weighted magnetization-prepared rapid acquisition gradient-echo sequence. Then, a gradient-echo T2*-weighted echo-planar imaging sequence was performed for VGT and HMT. Scan parameters are described in supplementary materials. 2.3. fMRI paradigms Two fMRI tasks were used: VGT and HMT. For the VGT, 12 blocks of alternating control and activation conditions were performed. After a 6-s fixation period, participants completed the 6-min task composed of 30-s blocks. During the activation condition, participants were asked to generate single verbs for concrete nouns presented visually every 3 seconds (ten nouns per block). During the control condition, participants were required to silently repeat two letters presented among four symbols. Participants practiced the task overtly beforehand but were instructed to respond silently inside the scanner to minimize motion artefacts associated with speech (Sanjuán et al., 2010 ). In HMT, we adapted the procedure used by Pujol et al. ( 1998 ) to activate the primary motor cortex. Patients were instructed to repeatedly flex and extend all the fingers of one hand and then the other at a frequency of approximately one cycle per second. Over a six-minute period, they alternated between moving the right and left hand in 30-second blocks, without any intermediate rest period between hands. Patients always began with the hand ipsilateral to the cerebral lesion. The instruction to switch hands was delivered verbally through an intercom two seconds before the end of each 30-second block. They were asked to maintain a constant rhythm and avoid moving their arms or head, and they practiced the motor task prior to entering the scanner. We chose this procedure to reliably and effectively obtain the activation of both hands separately. 2.4. fMRI task analysis TB data sets were processed by the Statistical Parametric Mapping software package (SPM12; Wellcome Trust Centre for Neuroimaging, London, UK) in native space. HMT pre-processing included: (1) alignment of images to AC-PC plane; (2) head motion correction, realignment and reslicing; and (3) spatial smooth (FWHM = 4mm). The GLM was applied by defining a boxcar function representing activation and control blocks convolved with the hemodynamic response function. Realignment parameters were added as nuisance variables and a high-pass (128s) filter was applied. Finally, one-sample t-tests were conducted to identify statistically significant active areas at p = .0001 (standard threshold routinely used by our group for presurgical motor mapping), which would serve as functional SMN. Pre-processing for the VGT data followed the same steps as previously described, with the addition of the following procedures between steps 2 and 3: co-registration of the T1-weighted structural images to the mean functional image; and segmentation of T1 structural images, applying an inverse deformation field. A normalized pipeline was used to determine group activation in both language and motor tasks including and to calculate language lateralization index (see Supplementary Materials) 2.5. scICA The Neuromark 2.1 template including 105 non-artifactual networks was used as a spatial constraint ( https://trendscenter.org/data/ ) (Iraji et al., 2022 ). The SMN spanned IC62 to IC74; however, only IC72 and IC73 fully encompassed the right and left central sulcus, respectively—from superior to inferior—including the characteristic omega-shaped hand knob. Therefore, these two components were extracted to represent the right and left SMN. Inverse deformation fields of VGT obtained from the segmentation step were applied to these templates for each patient. The resulting warped templates were spatially smoothed (FWHM = 4mm). Single-subject scICA was performed separately for each patient on VGT fMRI smoothed scans in native space by means of Group ICA of fMRI Toolbox (GIFT, Medical Image Analysis Lab, http://mialab.mrn.org/software/gift ). Number of IC was set on 2 and “Constrained ICA (Spatial)” was selected as the analysis algorithm. Then, default parameters were used. IC72 and IC73 maps, thresholded at z = 1.000, were combined for each patient and the resulting image was overlapped to each T1 scan. For each participant, a volume of interest was defined using MRIcron (Rorden & Brett, 2000 ) including central sulcus, precentral gyrus and postcentral gyrus, serving as predicted SMN. For each subject, we also confirmed that the hand activation maps were spatially aligned within the previously defined anatomical landmarks, which was the case in 40/40 (100%) participants. Unlike previous studies (Beheshtian et al., 2021 ), our method is fully objective and reproducible, as it relies on predefined anatomical boundaries rather than manual labelling or inter-rater agreement. 2.6. Statistical analysis For visualization only, group-level one-sample t-tests were performed for the language task (separating typical and atypical cases) and for the motor task (separating left and right activation) at voxel-wise p < .001 (uncorrected) and cluster-level family-wise error (FWE) corrected at p < .05. Individual functional SMN maps were obtained using single-subject one-sample t-tests from native HMT with a voxel-wise threshold of p < .001 (uncorrected), and cluster-level FWE corrected at p < .05, which corresponds to the standard threshold routinely used by our group for presurgical motor mapping. Performance metrics (sensitivity and specificity) were derived from these individual maps. Number of voxels of functional (HMT) and predicted SMN (IC72/IC73) and their overlap was calculated, taking active voxels in the HMT as the reference standard. True positive, true negative, false positive and false negative values were obtained for each patient. To determine true negative values, subject-specific masks generated from the pre-processed IC72 and IC73 templates were used as a reference landmark. Then, sensitivity (true positives / true positives + false negatives) and specificity (true negatives / true negatives + false positives) were calculated. Number of voxels was determined as the standard unit of measure for both methods and to conduct the sensitivity and specificity analysis. A sensitivity threshold of 50% was used to determine whether the predicted SMN sufficiently overlapped with the functional SMN. Predictions meeting or exceeding this threshold were classified as successfully localized. One factor ANOVA was performed to explore differences between sensitivity, and specificity values with sample’s variables. ANOVA results were corrected for multiple comparisons using the Benjamini-Hochberg False Discovery Rate (FDR) with q < .05. For completeness, uncorrected p-values are also shown to illustrate subthreshold trends. All statistical analyses were conducted by SPSS (IBM SPSS Statistics, v.28). 3. Results VGT showed group activation in the inferior frontal gyrus (IFG), including pars opercularis, and triangularis but also posterior areas (see Supplementary Materials). Activations were observed at voxel-wise p < .001 (uncorrected) and cluster-level FWE corrected at p < .05 in 38 patients in the left IFG, in 22 patients in right IFG, in 34 in the left temporal cortex and in 10 the right temporal cortex. LIs calculated in the IFG showed a mean of 45.65 (SD = 36.54) in the full sample. Typical left lateralization was observed for 29 patients (M = 64.07; SD = 9.415; range: 43 to 76), while 11 patients exhibited atypical language lateralization (M=-2.91; SD = 37.163; range: -59 to 39). This atypical lateralization was observed in one patient with left parietal lesion, nine patients with left frontal lesions, and one patient with left frontoparietal lesion. The HMT task revealed reliable group activations in the ipsilesional primary motor cortex, somatomotor cortex and supplementary motor area (voxel-wise p < .001 (uncorrected) and cluster-level FWE corrected at p activation). Eight patients (8/40, 20%) exhibited displacement of the central sulcus involving the SMN. In these patients, the displaced omega-shaped hand knob remained functionally active, suggesting no functional reorganization due to space-occupying lesion. ICA analysis scICA identified the SMN in all 40 patients, with each case reaching at least the predefined 50% sensitivity criterion. To determine the validity of this estimation of the sensorimotor area, we compared the activated voxels with those obtained with the native HMT ( p = .0001). Figure 1 summarizes the distribution of sensitivity and specificity values for both hemispheres, illustrating the variability and overall performance of the method. In the ipsilesional hemisphere, sensitivity values ranged from 50% to 100% and specificity values from 44.81% to 87.07%. In the contralesional hemisphere, sensitivity ranged from 50% to 99.48% and specificity from 61.48% to 92.65%. To provide a more comprehensive description of classification performance, additional confusion-matrix metrics, including number of voxels of predicted and functional SMN, true positives, true negatives, false positives, false negatives, sensitivity and specificity, are reported for each hemisphere and each patient in the supplementary materials. These metrics allow a patient-level evaluation of scICA-based SMN prediction relative to the HMT reference standard. Finally, paired-sample t-tests showed no significant differences between ipsilesional and contralesional sensitivity ( t (39) = 1.68, p = .10) or specificity ( t (39) = − 1.62, p = .11), indicating comparable performance of the method regardless of lesion side. In the 8 patients with displacement of the central sulcus, the mean sensitivity, and specificity of the SMN map extracted from scICA and compared to HMT were 80% (SD = 14%), and 70% (SD = 13%) in the ipsilateral hemisphere, and 77% (SD = 13%), and 79% (SD = 9%), in the contralateral hemisphere, respectively. Different ANOVAs were run to investigate the role of different acquisition and clinical variables (Table 2 for results). These demonstrated that 3T scanners produced higher ipsilesional and contralesional sensitivity than 1.5T scanners (Fig. 2 A). Also, contralesional sensitivity was higher in the right hemisphere than in the left, and for typical than atypical lateralization of language (Fig. 2 B and 2 C). Finally, ipsilesional specificity was higher when the central sulcus was not displaced by the tumor (Fig. 2 D). After applying Benjamini–Hochberg FDR correction to the 10 univariate ANOVAs per metric, only the effect of language lateralization on contralesional sensitivity remained significant ( p _FDR = .010). No other comparisons survived FDR correction. Table 2 F-values obtained in ANOVAs for variables type of lesion, lesion localization, lesion lobe, hemisphere, scanner, scanner (1.5T vs 3T), language lateralization, midline shift and ipsilesional central sulcus displacement. Variables Sensitivity Specificity Ipsilesional Hemisphere (left vs right) 1.27 (.27) .87 (.36) Scanner (1.5T vs 3T) 5.13 (.03)** .23 (.64) Language lateralization (typical vs. atypical) .16 (.69) 1.98 (.17) Midline shift (yes vs no) .007 (.93) .04 (.85) Central sulcus displacement (yes vs no) .31 (.58) 6.47 (.02)** Contralesional Hemisphere 4.57 (.04)** 2.13 (.15) Scanner (1.5T vs 3T) 4.31 (.05)** 3.68 (.06) Language lateralization (typical vs. atypical) 16.76 (< .001)* 1.66 (.21) Midline shift (yes vs no) 2.21 (.15) .46 (.50) Central sulcus displacement (yes vs no) .47 (.500) 1.03 (.32) Note. – Data are presented as Fisher’s F ( p value). Significant results surviving FDR correction are marked with * ( p < .001, FDR-corrected). Uncorrected effects are marked with ** ( p < .05, uncorrected). To illustrate individual results, Fig. 3 shows hand activation obtained from HMT at voxel-wise p < .0001 (uncorrected) and cluster-level FWE corrected at p < .05 overlapping predicted SMN extracted from scICA in four sample patients. A 3D video of results displayed in Fig. 3 can be seen in supplementary materials. 4. Discussion In this study, we introduce a new methodology that provides reliable presurgical assessments of language and motor functions using a 6-minute protocol. The data obtained in this sample using VGT have demonstrated high individual reliability in localizing the IFG, in line with findings previously reported in studies involving patient cohorts and healthy participants (Sanjuán et al., 2010 ; Villar-Rodríguez et al., 2020 ). Using scICA of this language task, we successfully localized the somatomotor network and primary motor cortex in all 40 patients (100%), including those exhibiting evident displacement of the central sulcus due to lesion-related mass effect (Beheshtian et al., 2021 ). In summary, we have obtained a reliable protocol for identifying language and motor areas in presurgical patients. Few previous studies take language tasks to predict SMN, and none of them employ quantitative measures to analyze the overlap between methods (Beheshtian et al., 2021 ; Jurkiewicz et al., 2018 ; Sparacia et al., 2019 ). Our sensitivity and specificity values were calculated at the voxel level by comparing the overlap between the SMN network obtained through scICA and the activation derived from the HMT. Using this protocol, voxel-wise sensitivity was good (78% ipsilesional; 74% contralesional), with 80% of cases showing good–excellent performance. Ipsilesional specificity averaged 75%, indicating low false positives, with only one outlier (44%). These results remained excellent even in the presence of mass effect, which displaced the sensorimotor areas. This opens the possibility of using structural T1-weighted images with anatomical masks and inverse deformations for rapid functional mapping. Future studies should explore this possibility. When comparing sensitivity with previous studies, one work did not report voxelwise sensitivity (Beheshtian et al., 2021 ), while another found lower sensitivity values (44%, Rosazza et al., 2014 ) when comparing RS and TB using ROIs or ICA. In contrast, using direct cortical stimulation as reference, both lower (63%; Qiu et al., 2017 ) and higher values (78–90%; Qiu et al., 2014 ) have been reported. However, those comparisons rely on non-equivalent units, and larger surface-based regions can inflate sensitivity estimates. Thus, our voxel-wise approach applies a stricter criterion, making results more comparable and robust. Although two previous studies have reported higher specificity values (84–89%, Qiu et al., 2014 ; and 93%, Qiu et al., 2017 ) when localizing motor areas, these results were obtained using direct cortical stimulation as the gold standard. While this technique is considered highly accurate, it differs fundamentally from our reference (fMRI functional task). As discussed previously, comparisons across studies must consider the variability introduced by the reference standard. In contrast, the study most comparable to ours —based on ICA and RS— reported considerably lower specificity values (27–51%, Rosazza et al., 2014 ). In this context, our mean specificity of 75% in the ipsilesional hemisphere demonstrates a substantial improvement in voxel-level spatial accuracy using scICA, supporting its clinical applicability in presurgical motor mapping. Some variables seem to modulate the results. As expected, higher magnetic field strength (3T) produces better results across all indices compared to a 1.5T scanner, possibly due to better spatial resolution and signal-to-noise ratio (Tyndall et al., 2017 ). As observed in previous studies (Lehéricy et al., 2000 ; Nitschke et al., 1998 ), ipsilesional sensitivity was slightly reduced when there was a displacement of the central sulcus as a consequence of the lesion, but the difference did not reach statistical significance, meaning that results for these cases were still optimal. Finally, contralesional hemisphere sensitivity was reduced in the left hemisphere and in patients with atypical language lateralization, an aspect that should be investigated in the future. 4.1. Limitations Some limitation should be addressed. First, thresholds of both TB and scICA results were set based on our previous research experience to demonstrate the utility of the technique, so further research may extend the analysis to compare different thresholds. Second, this study relies on connectivity analysis, which can differ from TB. However, both sensitivity and specificity of hand motor area have been previously validated for TB and RS using direct cortical stimulation with encouraging results (Qiu et al., 2014 ). Third, TB was used as the reference for sensitivity analysis, though it is not the gold standard (Rosazza et al., 2014 ); including cortical stimulation data would offer deeper insight. Finally, only hand motor task scans were available, so future research may include mouth and foot data. 5. Conclusion In conclusion, our study confirms the possibility of predicting somatosensory network connectivity from a language task and give valid information on language network distribution in patients diagnosed with space-occupying lesions. This method represents a promising alternative that can reduce scan duration and costs as only one fMRI sequence is needed to determine two different functional networks. Further research including direct cortical stimulation data or multiple threshold comparison may prove useful. Abbreviations RS = resting-state fMRI, TB = task-based fMRI, SMN = sensorimotor network, IC = independent component, scICA = spatially constrained independent component analysis, HMT = hand motor task, VGT = verb generation task, FWE= family-wise error. Declarations Declaration of Interest statement : authors have nothing to declare. Funding sources: This work was supported by the Spanish State Research Agency (Agencia Estatal de Investigación, AEI) [grant number CPP2023-010601, and PID2023-150776NB-I00, awarded to C. Ávila, and RYC2021-033809-I awarded to V. Costumero]; by the Directorate General for Science and Research of the Generalitat Valenciana [grant number CIPROM/2023/58, awarded to C. Ávila], and by NextGenerationEU funds through the Spanish Recovery, Transformation and Resilience Plan [NextGenerationEU/PRTR, awarded to V. Costumero]. Conflicts of interest: The authors declare that they have no conflicts of interest related to the conduct or reporting of this research. Data, material and/or code availability: All data supporting the findings of this study are available from the corresponding author on reasonable request, in accordance with institutional and ethical regulations. Ethics approval: This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. The research protocol was reviewed and approved by the Ethics Committee of the Universitat Jaume I (Castellón de la Plana, Spain). All participants provided written informed consent for the use of their clinical and neuroimaging data for research purposes (participation in the study and publication of anonymized data). Author’s contribution statement : Author contributions included conception and study design (EC-R, VC and CA), data collection or acquisition (EC-R, VC and CA), statistical analysis (EC-R, VC and VB), interpretation of results (all authors), drafting the manuscript work or revising it critically for important intellectual content (EC-R, VC and CA) and approval of final version to be published and agreement to be accountable for the integrity and accuracy of all aspects of the work (All authors). Clinical trial number: not applicable References Beheshtian, E., Jalilianhasanpour, R., Shanechi, A. M., Sethi, V., Wang, G., Lindquist, M. A., Caffo, B. S., Agarwal, S., Pillai, J. J., Gujar, S. K., & Sair, H. I. (2021). Identification of the somatomotor network from language task-based fMRI compared with resting-state fMRI in patients with brain lesions. Radiology , 301 (1), 178–184. https://doi.org/10.1148/radiol.2021204594 Branco, P., Seixas, D., Deprez, S., Kovacs, S., Peeters, R., Sl, C., & Sunaert, S. (2016). Resting-State Functional Magnetic Resonance Imaging for Language Preoperative Planning. Front. Hum. Neurosci , 10 , 11. https://doi.org/10.3389/fnhum.2016.00011 Buchbinder, B. R. (2016). Functional magnetic resonance imaging. In J. C. Masdeu & R. G. González (Eds.), Handbook of Clinical Neurology (3rd ed., Vol. 135, pp. 61–92). Elsevier B.V. https://doi.org/10.1016/B978-0-444-53485-9.00004-0 Dierker, D., Roland, J. L., Kamran, M., Rutlin, J., Hacker, C. D., Marcus, D. S., Milchenko, M., Miller-Thomas, M. M., Benzinger, T. L., Snyder, A. Z., Leuthardt, E. C., & Shimony, J. S. (2017). Resting-state Functional Magnetic Resonance Imaging in Presurgical Functional Mapping: Sensorimotor Localization. Neuroimaging Clinics of North America , 27 (4), 621–633. https://doi.org/10.1016/J.NIC.2017.06.011 Iraji, A., Fu, Z., Faghiri, A., Duda, M., Chen, J., Rachakonda, S., DeRamus, T., Kochunov, P., Adhikari, B. M., Belger, A., Ford, J. M., Mathalon, D. H., Pearlson, G. D., Potkin, S. G., Preda, A., Turner, J. A., Erp, T. G. M. van, Bustillo, J. R., Yang, K., … Calhoun, V. D. (2022). Canonical and Replicable Multi-Scale Intrinsic Connectivity Networks in 100k+ Resting-State fMRI Datasets. BioRxiv , 2022.09.03.506487. https://doi.org/10.1101/2022.09.03.506487 Jurkiewicz, M. T., Crawley, A. P., & Mikulis, D. J. (2018). Is rest really rest? resting-state functional connectivity during rest and motor task paradigms. Brain Connectivity , 8 (5), 268–275. https://doi.org/10.1089/brain.2017.0495 Lakhani, D. A., Sabsevitz, D. S., Chaichana, K. L., Quiñones-Hinojosa, A., & Middlebrooks, E. H. (2023). Current State of Functional MRI in the Presurgical Planning of Brain Tumors. Radiology: Imaging Cancer , 5 (6). https://doi.org/10.1148/RYCAN.230078/ASSET/IMAGES/LARGE/RYCAN.230078.FIG5.JPEG Lehéricy, S., Duffau, H., Cornu, P., Capelle, L., Pidoux, B., Carpentier, A., Auliac, S., Clemenceau, S., Sichez, J. P., Bitar, A., Valery, C. A., Van Effenterre, R., Faillot, T., Srour, A., Fohanno, D., Philippon, J., Le Bihan, D., & Marsault, C. (2000). Correspondence between functional magnetic resonance imaging somatotopy and individual brain anatomy of the central region: comparison with intraoperative stimulation in patients with brain tumors. Journal of Neurosurgery , 92 (4), 589–598. https://doi.org/10.3171/JNS.2000.92.4.0589 Lin, Q.-H., Liu, J., Zheng, Y.-R., Liang, H., & Calhoun, V. D. (2010). Semiblind Spatial ICA of fMRI Using Spatial Constraints. Human Brain Mapping , 31 , 1076–1088. https://doi.org/10.1002/hbm.20919 Lu, J., Zhang, H., Hameed, N. U. F., Zhang, J., Yuan, S., Qiu, T., Shen, D., Wu, J., & Lu, K. J. (2017). An automated method for identifying an independent component analysis-based language-related resting-state network in brain tumor subjects for surgical planning. Scientific Reports , 7 (13769). https://doi.org/10.1038/s41598-017-14248-5 Lu, W., & Rajapakse, J. C. (2005). Approach and applications of constrained ICA. IEEE Transactions on Neural Networks , 16 (1), 203–212. https://doi.org/10.1109/TNN.2004.836795 Nitschke, M. F., Melchert, U. H., Hahn, C., Otto, V., Arnold, H., Herrmann, H. D., Nowak, G., Westphal, M., & Wessel, K. (1998). Preoperative functional magnetic resonance imaging (fMRI) of the motor system in patients with tumours in the parietal lobe. Acta Neurochirurgica , 140 (12), 1223–1229. https://doi.org/10.1007/S007010050242 Ott, C., Rosengarth, K., Doenitz, C., Hoehne, J., Wendl, C., Dodoo-Schittko, F., Lang, E., Schmidt, N. O., & Goldhacker, M. (2021). Preoperative assessment of language dominance through combined resting-state and task-based functional magnetic resonance imaging. Journal of Personalized Medicine , 11 (12), 1–14. https://doi.org/10.3390/jpm11121342 Parker Jones, O., Voets, N. L., Adcock, J. E., Stacey, R., & Jbabdi, S. (2017). Resting connectivity predicts task activation in pre-surgical populations. NeuroImage: Clinical , 13 , 378–385. https://doi.org/10.1016/j.nicl.2016.12.028 Pujol, J., Conesa, G., Deus, J., López-Obarrio, L., Isamat, F., & Capdevila, A. (1998). Clinical application of functional magnetic resonance imaging in presurgical identification of the central sulcus. Journal of Neurosurgery , 88 (5), 863–869. https://doi.org/10.3171/JNS.1998.88.5.0863 Qiu, T. M., Gong, F. Y., Gong, X., Wu, J. S., Lin, C. P., Biswal, B. B., Zhuang, D. X., Yao, C. J., Zhang, X. L., Lu, J. F., Zhu, F. P., Mao, Y., & Zhou, L. F. (2017). Real-Time Motor Cortex Mapping for the Safe Resection of Glioma: An Intraoperative Resting-State fMRI Study. AJNR. American Journal of Neuroradiology , 38 (11), 2146–2152. https://doi.org/10.3174/AJNR.A5369 Qiu, T. ming, Yan, C. gan, Tang, W. jun, Wu, J. song, Zhuang, D. xiao, Yao, C. jun, Lu, J. feng, Zhu, F. ping, Mao, Y., & Zhou, L. fu. (2014). Localizing hand motor area using resting-state fMRI: validated with direct cortical stimulation. Acta Neurochirurgica , 156 (12), 2295–2302. https://doi.org/10.1007/S00701-014-2236-0/FIGURES/2 Rolinski, R., You, X., Gonzalez-Castillo, J., Norato, G., Reynolds, R. C., Inati, S. K., & Theodore, W. H. (2020). Language lateralization from task-based and resting state functional MRI in patients with epilepsy. Human Brain Mapping , 41 (11), 3133–3146. https://doi.org/10.1002/hbm.25003 Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions. Behavioural Neurology , 12 , 191–200. https://doi.org/10.1155/2000/421719 Rosazza, C., Aquino, D., ’incerti, D., Cordella, L., & Andronache, R. (2014). Preoperative Mapping of the Sensorimotor Cortex: Comparative Assessment of Task-Based and Resting-State fMRI. PLoS ONE , 9 (6), 98860. https://doi.org/10.1371/journal.pone.0098860 Sair, H. I., Agarwal, S., & Pillai, J. J. (2017). Application of Resting State Functional MR Imaging to Presurgical Mapping: Language Mapping. Neuroimaging Clinics of North America , 27 (4), 635–644. https://doi.org/10.1016/j.nic.2017.06.003 Sanjuán, A., Bustamante, J. C., Forn, C., Ventura-Campos, N., Barrós-Loscertales, A., Martínez, J. C., Villanueva, V., & Ávila, C. (2010). Comparison of two fMRI tasks for the evaluation of the expressive language function. Neuroradiology , 52 (5), 407–415. https://doi.org/10.1007/s00234-010-0667-8 Seitzman, B. A., Snyder, A. Z., Leuthardt, E. C., & Shimony, J. S. (2019). The State of Resting State Networks. Topics in Magnetic Resonance Imaging : TMRI , 28 (4), 189. https://doi.org/10.1097/RMR.0000000000000214 Sparacia, G., Parla, G., Cannella, R., Perri, A., Lo Re, V., Mamone, G., Miraglia, R., Torregrossa, F., & Grasso, G. (2019). Resting-State Functional Magnetic Resonance Imaging for Brain Tumor Surgical Planning: Feasibility in Clinical Setting. World Neurosurgery , 131 , 356–363. https://doi.org/10.1016/j.wneu.2019.07.022 Sparacia, G., Parla, G., Mamone, G., Caruso, M., Torregrossa, F., & Grasso, G. (2021). Resting-State Functional Magnetic Resonance Imaging for Surgical Neuro-Oncology Planning: Towards a Standardization in Clinical Settings. Brain Sciences 2021, Vol. 11, Page 1613 , 11 (12), 1613. https://doi.org/10.3390/BRAINSCI11121613 Tie, Y., Rigolo, L., Norton, I. H., Huang, R. Y., Wu, W., Orringer, D., Mukundan, S., & Golby, A. J. (2014). Defining Language Networks From Resting-State fMRI for Surgical Planning-A Feasibility Study. Human Brain Mapping , 35 , 1018–1030. https://doi.org/10.1002/hbm.22231 Tyndall, A. J., Reinhardt, J., Tronnier, V., Mariani, L., & Stippich, C. (2017). Presurgical motor, somatosensory and language fMRI: Technical feasibility and limitations in 491 patients over 13 years. European Radiology , 27 (1), 267–278. https://doi.org/10.1007/S00330-016-4369-4/TABLES/3 Villar‐Rodríguez, E., Palomar‐García, M., Hernández, M., Adrián‐Ventura, J., Olcina‐Sempere, G., Parcet, M., & Ávila, C. (2020). Left‐handed musicians show a higher probability of atypical cerebral dominance for language. Human Brain Mapping , 41 (8), 2048. https://doi.org/10.1002/HBM.24929 Additional Declarations No competing interests reported. Supplementary Files 5.SupplementalFile.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 08 May, 2026 Reviews received at journal 11 Apr, 2026 Reviewers agreed at journal 29 Mar, 2026 Reviewers invited by journal 18 Dec, 2025 Editor assigned by journal 16 Dec, 2025 Submission checks completed at journal 09 Dec, 2025 First submitted to journal 05 Dec, 2025 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-8287975","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":562459989,"identity":"d25d4ab3-d7ec-4fbf-9d0d-d41d83e7a91b","order_by":0,"name":"Eva Calderón-Rubio","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvklEQVRIiWNgGAWjYHACxgM8QJKfgSHhANF6wFokG0jWYkC0en72ww8OvKm4J2d8I+HhAYaKOsJaJHvSDA7OOVNsbHYjAeiwM4cJazG4wWBwmLctIXEbSAtjGxHOs7/B/uEw77+ExM0zQFr+EeEwAwkeoC0NCYkbJEBaGpgJa5E4k1NwcM6xBGOJMw8SDiQcI8Iv/O3HNz54U5Mgx9+ek/zhQw0RDkMCPAkMCSRpYGBgP0CihlEwCkbBKBgpAAANv0ETfZMKSgAAAABJRU5ErkJggg==","orcid":"","institution":"Jaume I University","correspondingAuthor":true,"prefix":"","firstName":"Eva","middleName":"","lastName":"Calderón-Rubio","suffix":""},{"id":562459990,"identity":"f4e7608b-d7fe-4d93-aa0f-2e8435708787","order_by":1,"name":"Víctor Costumero","email":"","orcid":"","institution":"Jaume I University","correspondingAuthor":false,"prefix":"","firstName":"Víctor","middleName":"","lastName":"Costumero","suffix":""},{"id":562459991,"identity":"3d43b56b-f43f-4e84-99a1-11769d614252","order_by":2,"name":"Vicente Belloch","email":"","orcid":"","institution":"Ascires Grupo Biomédico","correspondingAuthor":false,"prefix":"","firstName":"Vicente","middleName":"","lastName":"Belloch","suffix":""},{"id":562459993,"identity":"738b80ff-fd18-4eef-8992-2ceab25f0456","order_by":3,"name":"César Ávila","email":"","orcid":"","institution":"Jaume I University","correspondingAuthor":false,"prefix":"","firstName":"César","middleName":"","lastName":"Ávila","suffix":""}],"badges":[],"createdAt":"2025-12-05 12:53:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8287975/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8287975/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99307949,"identity":"24d55455-3ae1-424c-bd25-d55714d06d65","added_by":"auto","created_at":"2025-12-31 16:07:11","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":572994,"visible":true,"origin":"","legend":"","description":"","filename":"3.Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/732db25e7587ca545d291c79.docx"},{"id":99307813,"identity":"ee273abb-9d7f-4457-9b5e-c0a124e58fd1","added_by":"auto","created_at":"2025-12-31 16:06:51","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6444,"visible":true,"origin":"","legend":"","description":"","filename":"6859fbcfb23b4777bc6c29ddb09a413b.json","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/b9e4f83dd4dc40f98c04637f.json"},{"id":99307760,"identity":"998a112e-6188-4714-b3e2-ee4ea5b223c5","added_by":"auto","created_at":"2025-12-31 16:06:42","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":646517,"visible":true,"origin":"","legend":"","description":"","filename":"5.SupplementalFile.docx","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/4bf6c4a38f819f774b4307c1.docx"},{"id":99307458,"identity":"f5d3c698-da3b-4440-9fe0-260d1c25bf50","added_by":"auto","created_at":"2025-12-31 16:06:17","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":111797,"visible":true,"origin":"","legend":"","description":"","filename":"6859fbcfb23b4777bc6c29ddb09a413b1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/845ccbeaf3884e3aaf9e1799.xml"},{"id":99307931,"identity":"4bd9bf4c-df26-492d-a75d-68834334bf22","added_by":"auto","created_at":"2025-12-31 16:07:06","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58220,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/c802459f58e44cb6f7c810ec.jpeg"},{"id":99307336,"identity":"acbe6cfe-2c88-40bf-b139-015ec1491e09","added_by":"auto","created_at":"2025-12-31 16:06:03","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":55906,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/e083f3d30cac9318377ecc89.jpeg"},{"id":99307302,"identity":"e5110ec1-4fc2-499f-a280-39c9361d8ab6","added_by":"auto","created_at":"2025-12-31 16:05:56","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":229013,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/78f6f4a7f9c46ab7d0270d3f.jpeg"},{"id":98820867,"identity":"e2049c80-ef1a-4d9e-89f5-ab27c736dfa2","added_by":"auto","created_at":"2025-12-22 17:19:00","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10667,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/23c95c14e2c9b9cd62b6ed0f.png"},{"id":99307471,"identity":"8f586723-d4e7-485d-b96e-6d2ba8ea445d","added_by":"auto","created_at":"2025-12-31 16:06:19","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":11198,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/c1485ea6487df6b390bad071.png"},{"id":99307200,"identity":"30196d9a-c6c8-4d70-a3fd-363e8b14a264","added_by":"auto","created_at":"2025-12-31 16:05:47","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106368,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/59c92f47a29f6f07c2b9603e.png"},{"id":98820878,"identity":"d50dd917-e99b-4238-8992-621295d8c077","added_by":"auto","created_at":"2025-12-22 17:19:01","extension":"xml","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":110238,"visible":true,"origin":"","legend":"","description":"","filename":"6859fbcfb23b4777bc6c29ddb09a413b1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/331deafc42335ab4d7f5a2eb.xml"},{"id":99307856,"identity":"e5010bce-7732-409c-8b05-ca614bde747c","added_by":"auto","created_at":"2025-12-31 16:06:55","extension":"html","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":123417,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/a552d17e85c06add0383771b.html"},{"id":99307454,"identity":"84d062ec-b10e-4a2e-ad0f-6e64be66627c","added_by":"auto","created_at":"2025-12-31 16:06:17","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":27631,"visible":true,"origin":"","legend":"\u003cp\u003eBar diagram showing average sample results of sensitivity, and specificity for both ipsilesional (dark-grey) and contralesional (light-grey) hemispheres. Mean percentage (standard deviation).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/21b0693c1c7eb09843f47c10.png"},{"id":99307626,"identity":"a5c2705f-f25c-43b6-99ca-7345c9550cda","added_by":"auto","created_at":"2025-12-31 16:06:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":32549,"visible":true,"origin":"","legend":"\u003cp\u003eBar diagram of ANOVA differences in statistically significant variables. The X-axis depicts the different variables, while the Y-axis shows the mean outcome expressed as a percentage. Mean percentage (standard deviation). a) Scanner differences in ipsilesional and contralesional sensitivity. Dark grey: 3T, and light grey: 1.5T scanner. b) Right and left lesion hemisphere differences in contralesional sensitivity. Dark grey: right lesion, and light grey: left lesion. C) Typical and atypical language lateralization differences in contralesional sensitivity. Dark grey: typical language lateralization, and light grey: atypical language lateralization. d) Displacement and not displacement of central sulcus differences in ipsilesional specificity. Dark grey: not displaced central sulcus, and light grey: displaced central sulcus. Significant results surviving FDR correction are marked with * (\u003cem\u003ep\u003c/em\u003e\u0026lt; .001, FDR-corrected). Uncorrected effects are marked with ** (\u003cem\u003ep\u003c/em\u003e \u0026lt; .05, uncorrected).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/d3a9ee505f48a6c715623c74.png"},{"id":99307305,"identity":"c234caf0-904c-48c6-9e8c-37c977b693f8","added_by":"auto","created_at":"2025-12-31 16:05:57","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":218062,"visible":true,"origin":"","legend":"\u003cp\u003eOverlap of hand motor task and SMN extracted from scICA. Green overlay: sensorimotor network obtained from volume of interest of the combination of IC72 and IC73 of each patient’s scICA; red overlays: hand functional activation extracted from each patient’s hand motor task in native space; yellow overlays: overlapping regions. a) 32-years-old man with left frontal diffuse astrocytoma. Sensitivity and specificity (%) for the ipsilesional hemisphere were 90.00 and 79.99, respectively; for the contralesional hemisphere, 86.47 and 81.87, respectively. b) 23-years-old woman with frontoparietal hemangiopericytoma. Sensitivity and specificity (%) for the ipsilesional hemisphere were 100.00 and 79.81, respectively; for the contralesional hemisphere, 62.02 and 84.97, respectively. c) 33-years-old man with left frontal diffuse astrocytoma. Sensitivity and specificity (%) for the ipsilesional hemisphere were 95.08 and 77.50, respectively; for the contralesional hemisphere, 80.65 and 77.94, respectively. d) 23-years-old woman with left frontal oligodendroglioma. Sensitivity and specificity (%) for the ipsilesional hemisphere were 89.72 and 79.95, respectively; for the contralesional hemisphere, 94.80 and 77.92, respectively. scICA=spatially constrained independent component analysis.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/f60c8c62e24ff450344f9e80.png"},{"id":99322220,"identity":"2f07cad2-579b-40ab-b335-38fc66607ae1","added_by":"auto","created_at":"2025-12-31 16:43:12","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1011779,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/26a6fecf-5ed7-4078-9efa-6004678aea98.pdf"},{"id":99307420,"identity":"f5d4693e-4b7f-44f8-8304-f00ee369d03c","added_by":"auto","created_at":"2025-12-31 16:06:14","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":646517,"visible":true,"origin":"","legend":"","description":"","filename":"5.SupplementalFile.docx","url":"https://assets-eu.researchsquare.com/files/rs-8287975/v1/31747633039c7e96ab4ab506.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eSpatially constrained ICA from a language task predicts Sensorimotor Network in patients with tumors\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSurgical resection of brain lesions requires balancing maximal tumor removal with the preservation of cognitive and sensorimotor functions, aiming to avoid permanent postoperative deficits (Lakhani et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). fMRI is a widely used non-invasive tool for presurgical planning, particularly for mapping language and motor functions (Buchbinder, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). However, task-based fMRI (TB) presents limitations: results depend heavily on patient performance, posing challenges for individuals with cognitive impairment or claustrophobia, and the procedure is costly, time-consuming, and demands active patient cooperation (Leuthardt et al., 2015; Pujol et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e1998\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo address the limitations of TB, resting-state (RS) paradigms have emerged as a promising alternative (Parker Jones et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Sparacia et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). RS offers several advantages: it can reveal multiple functional networks beyond language and motor systems (Seitzman et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), and it does not require active patient participation (Dierker et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). While RS has shown encouraging results for motor mapping (Qiu et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), evidence for language mapping remains limited and less consistent (Sair et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). RS often shows weaker language lateralization and is less sensitive in cases with atypical lateralization, which may lead to misleading presurgical decisions (Branco et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; J. Lu et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Ott et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rolinski et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Sparacia et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Tie et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eGiven the difficulty of using RS to determine language lateralization, an alternative is to use a single language task and apply ICA to identify SMN (Beheshtian et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Jurkiewicz et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, limitations remain: Jurkiewicz et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) performs a seed-based analysis of normalized data in a healthy subject sample that can be limiting when working in native space of presurgical patients diagnosed with space-occupying lesions; and Beheshtian et al. (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) did not detect SMN independent components (IC) in all participants and reported RS\u0026ndash;TB correspondence in only 75% of cases.\u003c/p\u003e \u003cp\u003eOur objective in this study is to overcome the above-mentioned limitations, and to do so we propose a novel methodology that uses a language task-based paradigm to pursue three goals: (1) identify language lateralization; (2) predict SMN activation from a language task; and (3) assess the sensitivity and specificity of these results in comparison with a motor fMRI task (Qiu et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Qiu et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Rosazza et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). To address the shortcomings of previous approaches, we apply spatially constrained ICA (scICA) to language-task preprocessed scans. scICA automatically extracts the ICs of interest using a predefined spatial template that serves as a constraint on the source matrix (Lin et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Lu \u0026amp; Rajapakse, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). Our main hypothesis is that a language task can be used not only for presurgical language mapping but also to accurately map motor areas of the brain using ICA.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Participants.\u003c/h2\u003e \u003cp\u003eA group of 40 patients (21 males, ages 17\u0026ndash;78 years, 40.50\u0026thinsp;\u0026plusmn;\u0026thinsp;13.99) was selected retrospectively (2009\u0026ndash;2022) from a pool of 288 cases for meeting the following inclusion criteria: (1) a space-occupying brain lesion candidate for neurosurgery; (2) a motor task for hand movement (HMT); and (3) a verb-generation task (VGT). Exclusion criteria were incomplete tasks, motion artifacts, or poor in-scanner performance.\u003c/p\u003e \u003cp\u003eDemographic data, pathology information, and other variables of interest are reported in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. Prior to scanning, all patients signed an informed consent form, providing written consent to participate in the study. The study protocol was approved by the university Ethics Committee.\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\u003e\u003cb\u003eDemographic data, tumor location, pathologic anatomy analysis results, language lateralization (LI), and type of scanner.\u003c/b\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eParameters\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMidline shift\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eIpsilesional central sulcus displacement\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLesion location\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eHemisphere: left / right\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e29 / 11\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eFrontal lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eParietal lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eTemporoparietal lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eFrontotemporal lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eFrontoparietal lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eAnatomic pathology result\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eOligodendroglioma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eGlioma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMultiform glioblastoma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eAnaplastic astrocytoma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eDiffuse astrocytoma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eOligoastrocytoma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMetastatic lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMeningioma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eCavernoma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eCavernous hemangioma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eHemangiopericytoma\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eDNET\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eUndetermined Space-occupying brain lesion\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eScanner\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eSiemens TrioTim 3T\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eSiemens Sonata 1.5T\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003ePhilips Achieva 3T\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003eNote: values are number of patients.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Image Acquisition\u003c/h2\u003e \u003cp\u003eThree different scanners were used: Siemens Magnetom Trio 3T, Siemens Magnetom Sonata 1.5T and Philips Achieva 3T X-series. First, a 3D structural MRI was acquired using a T1-weighted magnetization-prepared rapid acquisition gradient-echo sequence. Then, a gradient-echo T2*-weighted echo-planar imaging sequence was performed for VGT and HMT. Scan parameters are described in supplementary materials.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. fMRI paradigms\u003c/h2\u003e \u003cp\u003eTwo fMRI tasks were used: VGT and HMT. For the VGT, 12 blocks of alternating control and activation conditions were performed. After a 6-s fixation period, participants completed the 6-min task composed of 30-s blocks. During the activation condition, participants were asked to generate single verbs for concrete nouns presented visually every 3 seconds (ten nouns per block). During the control condition, participants were required to silently repeat two letters presented among four symbols. Participants practiced the task overtly beforehand but were instructed to respond silently inside the scanner to minimize motion artefacts associated with speech (Sanju\u0026aacute;n et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). In HMT, we adapted the procedure used by Pujol et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e1998\u003c/span\u003e) to activate the primary motor cortex. Patients were instructed to repeatedly flex and extend all the fingers of one hand and then the other at a frequency of approximately one cycle per second. Over a six-minute period, they alternated between moving the right and left hand in 30-second blocks, without any intermediate rest period between hands. Patients always began with the hand ipsilateral to the cerebral lesion. The instruction to switch hands was delivered verbally through an intercom two seconds before the end of each 30-second block. They were asked to maintain a constant rhythm and avoid moving their arms or head, and they practiced the motor task prior to entering the scanner. We chose this procedure to reliably and effectively obtain the activation of both hands separately.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. fMRI task analysis\u003c/h2\u003e \u003cp\u003eTB data sets were processed by the Statistical Parametric Mapping software package (SPM12; Wellcome Trust Centre for Neuroimaging, London, UK) in native space.\u003c/p\u003e \u003cp\u003eHMT pre-processing included: (1) alignment of images to AC-PC plane; (2) head motion correction, realignment and reslicing; and (3) spatial smooth (FWHM\u0026thinsp;=\u0026thinsp;4mm). The GLM was applied by defining a boxcar function representing activation and control blocks convolved with the hemodynamic response function. Realignment parameters were added as nuisance variables and a high-pass (128s) filter was applied. Finally, one-sample t-tests were conducted to identify statistically significant active areas at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.0001 (standard threshold routinely used by our group for presurgical motor mapping), which would serve as functional SMN.\u003c/p\u003e \u003cp\u003ePre-processing for the VGT data followed the same steps as previously described, with the addition of the following procedures between steps 2 and 3: co-registration of the T1-weighted structural images to the mean functional image; and segmentation of T1 structural images, applying an inverse deformation field. A normalized pipeline was used to determine group activation in both language and motor tasks including and to calculate language lateralization index (see Supplementary Materials)\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. scICA\u003c/h2\u003e \u003cp\u003eThe Neuromark 2.1 template including 105 non-artifactual networks was used as a spatial constraint (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://trendscenter.org/data/\u003c/span\u003e\u003cspan address=\"https://trendscenter.org/data/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) (Iraji et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The SMN spanned IC62 to IC74; however, only IC72 and IC73 fully encompassed the right and left central sulcus, respectively\u0026mdash;from superior to inferior\u0026mdash;including the characteristic omega-shaped hand knob. Therefore, these two components were extracted to represent the right and left SMN. Inverse deformation fields of VGT obtained from the segmentation step were applied to these templates for each patient. The resulting warped templates were spatially smoothed (FWHM\u0026thinsp;=\u0026thinsp;4mm).\u003c/p\u003e \u003cp\u003eSingle-subject scICA was performed separately for each patient on VGT fMRI smoothed scans in native space by means of Group ICA of fMRI Toolbox (GIFT, Medical Image Analysis Lab, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://mialab.mrn.org/software/gift\u003c/span\u003e\u003cspan address=\"http://mialab.mrn.org/software/gift\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Number of IC was set on 2 and \u0026ldquo;Constrained ICA (Spatial)\u0026rdquo; was selected as the analysis algorithm. Then, default parameters were used.\u003c/p\u003e \u003cp\u003eIC72 and IC73 maps, thresholded at z\u0026thinsp;=\u0026thinsp;1.000, were combined for each patient and the resulting image was overlapped to each T1 scan. For each participant, a volume of interest was defined using MRIcron (Rorden \u0026amp; Brett, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) including central sulcus, precentral gyrus and postcentral gyrus, serving as predicted SMN. For each subject, we also confirmed that the hand activation maps were spatially aligned within the previously defined anatomical landmarks, which was the case in 40/40 (100%) participants. Unlike previous studies (Beheshtian et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), our method is fully objective and reproducible, as it relies on predefined anatomical boundaries rather than manual labelling or inter-rater agreement.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6. Statistical analysis\u003c/h2\u003e \u003cp\u003eFor visualization only, group-level one-sample t-tests were performed for the language task (separating typical and atypical cases) and for the motor task (separating left and right activation) at voxel-wise \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001 (uncorrected) and cluster-level family-wise error (FWE) corrected at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05.\u003c/p\u003e \u003cp\u003eIndividual functional SMN maps were obtained using single-subject one-sample t-tests from native HMT with a voxel-wise threshold of \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001 (uncorrected), and cluster-level FWE corrected at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05, which corresponds to the standard threshold routinely used by our group for presurgical motor mapping. Performance metrics (sensitivity and specificity) were derived from these individual maps.\u003c/p\u003e \u003cp\u003eNumber of voxels of functional (HMT) and predicted SMN (IC72/IC73) and their overlap was calculated, taking active voxels in the HMT as the reference standard. True positive, true negative, false positive and false negative values were obtained for each patient. To determine true negative values, subject-specific masks generated from the pre-processed IC72 and IC73 templates were used as a reference landmark. Then, sensitivity (true positives / true positives\u0026thinsp;+\u0026thinsp;false negatives) and specificity (true negatives / true negatives\u0026thinsp;+\u0026thinsp;false positives) were calculated. Number of voxels was determined as the standard unit of measure for both methods and to conduct the sensitivity and specificity analysis. A sensitivity threshold of 50% was used to determine whether the predicted SMN sufficiently overlapped with the functional SMN. Predictions meeting or exceeding this threshold were classified as successfully localized.\u003c/p\u003e \u003cp\u003eOne factor ANOVA was performed to explore differences between sensitivity, and specificity values with sample\u0026rsquo;s variables. ANOVA results were corrected for multiple comparisons using the Benjamini-Hochberg False Discovery Rate (FDR) with \u003cem\u003eq\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05. For completeness, uncorrected p-values are also shown to illustrate subthreshold trends. All statistical analyses were conducted by SPSS (IBM SPSS Statistics, v.28).\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003eVGT showed group activation in the inferior frontal gyrus (IFG), including pars opercularis, and triangularis but also posterior areas (see Supplementary Materials). Activations were observed at voxel-wise \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001 (uncorrected) and cluster-level FWE corrected at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05 in 38 patients in the left IFG, in 22 patients in right IFG, in 34 in the left temporal cortex and in 10 the right temporal cortex. LIs calculated in the IFG showed a mean of 45.65 (SD\u0026thinsp;=\u0026thinsp;36.54) in the full sample. Typical left lateralization was observed for 29 patients (M\u0026thinsp;=\u0026thinsp;64.07; SD\u0026thinsp;=\u0026thinsp;9.415; range: 43 to 76), while 11 patients exhibited atypical language lateralization (M=-2.91; SD\u0026thinsp;=\u0026thinsp;37.163; range: -59 to 39). This atypical lateralization was observed in one patient with left parietal lesion, nine patients with left frontal lesions, and one patient with left frontoparietal lesion.\u003c/p\u003e \u003cp\u003eThe HMT task revealed reliable group activations in the ipsilesional primary motor cortex, somatomotor cortex and supplementary motor area (voxel-wise \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001 (uncorrected) and cluster-level FWE corrected at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05; see Supplementary Materials). Our procedure also obtained significant activations of contralesional motor area in 100% of patients when applicated the opposite contrast (control\u0026thinsp;\u0026gt;\u0026thinsp;activation). Eight patients (8/40, 20%) exhibited displacement of the central sulcus involving the SMN. In these patients, the displaced omega-shaped hand knob remained functionally active, suggesting no functional reorganization due to space-occupying lesion.\u003c/p\u003e \u003cp\u003eICA analysis\u003c/p\u003e \u003cp\u003escICA identified the SMN in all 40 patients, with each case reaching at least the predefined 50% sensitivity criterion. To determine the validity of this estimation of the sensorimotor area, we compared the activated voxels with those obtained with the native HMT (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.0001). Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the distribution of sensitivity and specificity values for both hemispheres, illustrating the variability and overall performance of the method. In the ipsilesional hemisphere, sensitivity values ranged from 50% to 100% and specificity values from 44.81% to 87.07%. In the contralesional hemisphere, sensitivity ranged from 50% to 99.48% and specificity from 61.48% to 92.65%.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo provide a more comprehensive description of classification performance, additional confusion-matrix metrics, including number of voxels of predicted and functional SMN, true positives, true negatives, false positives, false negatives, sensitivity and specificity, are reported for each hemisphere and each patient in the supplementary materials. These metrics allow a patient-level evaluation of scICA-based SMN prediction relative to the HMT reference standard. Finally, paired-sample t-tests showed no significant differences between ipsilesional and contralesional sensitivity (\u003cem\u003et\u003c/em\u003e(39)\u0026thinsp;=\u0026thinsp;1.68, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.10) or specificity (\u003cem\u003et\u003c/em\u003e(39)\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;1.62, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;.11), indicating comparable performance of the method regardless of lesion side.\u003c/p\u003e \u003cp\u003eIn the 8 patients with displacement of the central sulcus, the mean sensitivity, and specificity of the SMN map extracted from scICA and compared to HMT were 80% (SD\u0026thinsp;=\u0026thinsp;14%), and 70% (SD\u0026thinsp;=\u0026thinsp;13%) in the ipsilateral hemisphere, and 77% (SD\u0026thinsp;=\u0026thinsp;13%), and 79% (SD\u0026thinsp;=\u0026thinsp;9%), in the contralateral hemisphere, respectively.\u003c/p\u003e \u003cp\u003eDifferent ANOVAs were run to investigate the role of different acquisition and clinical variables (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e for results). These demonstrated that 3T scanners produced higher ipsilesional and contralesional sensitivity than 1.5T scanners (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Also, contralesional sensitivity was higher in the right hemisphere than in the left, and for typical than atypical lateralization of language (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB and \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). Finally, ipsilesional specificity was higher when the central sulcus was not displaced by the tumor (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). After applying Benjamini\u0026ndash;Hochberg FDR correction to the 10 univariate ANOVAs per metric, only the effect of language lateralization on contralesional sensitivity remained significant (\u003cem\u003ep\u003c/em\u003e_FDR\u0026thinsp;=\u0026thinsp;.010). No other comparisons survived FDR correction.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eF-values obtained in ANOVAs for variables type of lesion, lesion localization, lesion lobe, hemisphere, scanner, scanner (1.5T vs 3T), language lateralization, midline shift and ipsilesional central sulcus displacement.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSensitivity\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSpecificity\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIpsilesional\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eHemisphere (left vs right)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.27 (.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.87 (.36)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eScanner (1.5T vs 3T)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5.13 (.03)**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.23 (.64)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eLanguage lateralization (typical vs. atypical)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e.16 (.69)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.98 (.17)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMidline shift (yes vs no)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e.007 (.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.04 (.85)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eCentral sulcus displacement (yes vs no)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e.31 (.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.47 (.02)**\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eContralesional\u003c/b\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 \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eHemisphere\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.57 (.04)**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.13 (.15)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eScanner (1.5T vs 3T)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.31 (.05)**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.68 (.06)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eLanguage lateralization (typical vs. atypical)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16.76 (\u0026lt;\u0026thinsp;.001)*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.66 (.21)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eMidline shift (yes vs no)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.21 (.15)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.46 (.50)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eCentral sulcus displacement (yes vs no)\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e.47 (.500)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.03 (.32)\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\u003e \u003c/p\u003e \u003cp\u003eNote. \u0026ndash; Data are presented as Fisher\u0026rsquo;s \u003cem\u003eF\u003c/em\u003e (\u003cem\u003ep\u003c/em\u003e value). Significant results surviving FDR correction are marked with * (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001, FDR-corrected). Uncorrected effects are marked with ** (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05, uncorrected).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo illustrate individual results, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows hand activation obtained from HMT at voxel-wise \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.0001 (uncorrected) and cluster-level FWE corrected at \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.05 overlapping predicted SMN extracted from scICA in four sample patients. A 3D video of results displayed in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e can be seen in supplementary materials.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn this study, we introduce a new methodology that provides reliable presurgical assessments of language and motor functions using a 6-minute protocol. The data obtained in this sample using VGT have demonstrated high individual reliability in localizing the IFG, in line with findings previously reported in studies involving patient cohorts and healthy participants (Sanju\u0026aacute;n et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Villar-Rodr\u0026iacute;guez et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Using scICA of this language task, we successfully localized the somatomotor network and primary motor cortex in all 40 patients (100%), including those exhibiting evident displacement of the central sulcus due to lesion-related mass effect (Beheshtian et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In summary, we have obtained a reliable protocol for identifying language and motor areas in presurgical patients.\u003c/p\u003e \u003cp\u003eFew previous studies take language tasks to predict SMN, and none of them employ quantitative measures to analyze the overlap between methods (Beheshtian et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Jurkiewicz et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Sparacia et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Our sensitivity and specificity values were calculated at the voxel level by comparing the overlap between the SMN network obtained through scICA and the activation derived from the HMT. Using this protocol, voxel-wise sensitivity was good (78% ipsilesional; 74% contralesional), with 80% of cases showing good\u0026ndash;excellent performance. Ipsilesional specificity averaged 75%, indicating low false positives, with only one outlier (44%). These results remained excellent even in the presence of mass effect, which displaced the sensorimotor areas. This opens the possibility of using structural T1-weighted images with anatomical masks and inverse deformations for rapid functional mapping. Future studies should explore this possibility.\u003c/p\u003e \u003cp\u003eWhen comparing sensitivity with previous studies, one work did not report voxelwise sensitivity (Beheshtian et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), while another found lower sensitivity values (44%, Rosazza et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) when comparing RS and TB using ROIs or ICA. In contrast, using direct cortical stimulation as reference, both lower (63%; Qiu et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) and higher values (78\u0026ndash;90%; Qiu et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) have been reported. However, those comparisons rely on non-equivalent units, and larger surface-based regions can inflate sensitivity estimates. Thus, our voxel-wise approach applies a stricter criterion, making results more comparable and robust.\u003c/p\u003e \u003cp\u003eAlthough two previous studies have reported higher specificity values (84\u0026ndash;89%, Qiu et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; and 93%, Qiu et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) when localizing motor areas, these results were obtained using direct cortical stimulation as the gold standard. While this technique is considered highly accurate, it differs fundamentally from our reference (fMRI functional task). As discussed previously, comparisons across studies must consider the variability introduced by the reference standard. In contrast, the study most comparable to ours \u0026mdash;based on ICA and RS\u0026mdash; reported considerably lower specificity values (27\u0026ndash;51%, Rosazza et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). In this context, our mean specificity of 75% in the ipsilesional hemisphere demonstrates a substantial improvement in voxel-level spatial accuracy using scICA, supporting its clinical applicability in presurgical motor mapping.\u003c/p\u003e \u003cp\u003eSome variables seem to modulate the results. As expected, higher magnetic field strength (3T) produces better results across all indices compared to a 1.5T scanner, possibly due to better spatial resolution and signal-to-noise ratio (Tyndall et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). As observed in previous studies (Leh\u0026eacute;ricy et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Nitschke et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1998\u003c/span\u003e), ipsilesional sensitivity was slightly reduced when there was a displacement of the central sulcus as a consequence of the lesion, but the difference did not reach statistical significance, meaning that results for these cases were still optimal. Finally, contralesional hemisphere sensitivity was reduced in the left hemisphere and in patients with atypical language lateralization, an aspect that should be investigated in the future.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Limitations\u003c/h2\u003e \u003cp\u003eSome limitation should be addressed. First, thresholds of both TB and scICA results were set based on our previous research experience to demonstrate the utility of the technique, so further research may extend the analysis to compare different thresholds. Second, this study relies on connectivity analysis, which can differ from TB. However, both sensitivity and specificity of hand motor area have been previously validated for TB and RS using direct cortical stimulation with encouraging results (Qiu et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Third, TB was used as the reference for sensitivity analysis, though it is not the gold standard (Rosazza et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2014\u003c/span\u003e); including cortical stimulation data would offer deeper insight. Finally, only hand motor task scans were available, so future research may include mouth and foot data.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn conclusion, our study confirms the possibility of predicting somatosensory network connectivity from a language task and give valid information on language network distribution in patients diagnosed with space-occupying lesions. This method represents a promising alternative that can reduce scan duration and costs as only one fMRI sequence is needed to determine two different functional networks. Further research including direct cortical stimulation data or multiple threshold comparison may prove useful.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eRS = resting-state fMRI, TB = task-based fMRI, SMN = sensorimotor network, IC = independent component, scICA = spatially constrained independent component analysis, HMT = hand motor task, VGT = verb generation task, FWE= family-wise error.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eDeclaration of Interest statement\u003c/strong\u003e: authors have nothing to declare.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding sources:\u0026nbsp;\u003c/strong\u003eThis work was supported by the Spanish State Research Agency (Agencia Estatal de Investigación, AEI) [grant number CPP2023-010601, and PID2023-150776NB-I00, awarded to C. Ávila, and RYC2021-033809-I awarded to V. Costumero]; by the Directorate General for Science and Research of the Generalitat Valenciana [grant number CIPROM/2023/58, awarded to C. Ávila], and by NextGenerationEU funds through the Spanish Recovery, Transformation and Resilience Plan [NextGenerationEU/PRTR, awarded to V. Costumero].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of interest:\u003c/strong\u003e The authors declare that they have no conflicts of interest related to the conduct or reporting of this research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData, material and/or code availability:\u003c/strong\u003e All data supporting the findings of this study are available from the corresponding author on reasonable request, in accordance with institutional and ethical regulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval:\u003c/strong\u003e This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. The research protocol was reviewed and approved by the Ethics Committee of the Universitat Jaume I (Castellón de la Plana, Spain). All participants provided written informed consent for the use of their clinical and neuroimaging data for research purposes (participation in the study and publication of anonymized data).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor’s contribution statement\u003c/strong\u003e: Author contributions included conception and study design (EC-R, VC and CA), data collection or acquisition (EC-R, VC and CA), statistical analysis (EC-R, VC and VB), interpretation of results (all authors), drafting the manuscript work or revising it critically for important intellectual content (EC-R, VC and CA) and approval of final version to be published and agreement to be accountable for the integrity and accuracy of all aspects of the work (All authors).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number:\u003c/strong\u003e not applicable\u003cbr\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBeheshtian, E., Jalilianhasanpour, R., Shanechi, A. M., Sethi, V., Wang, G., Lindquist, M. A., Caffo, B. S., Agarwal, S., Pillai, J. J., Gujar, S. K., \u0026amp; Sair, H. I. (2021). Identification of the somatomotor network from language task-based fMRI compared with resting-state fMRI in patients with brain lesions. \u003cem\u003eRadiology\u003c/em\u003e, \u003cem\u003e301\u003c/em\u003e(1), 178\u0026ndash;184. https://doi.org/10.1148/radiol.2021204594\u003c/li\u003e\n\u003cli\u003eBranco, P., Seixas, D., Deprez, S., Kovacs, S., Peeters, R., Sl, C., \u0026amp; Sunaert, S. (2016). Resting-State Functional Magnetic Resonance Imaging for Language Preoperative Planning. \u003cem\u003eFront. Hum. Neurosci\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 11. https://doi.org/10.3389/fnhum.2016.00011\u003c/li\u003e\n\u003cli\u003eBuchbinder, B. R. (2016). Functional magnetic resonance imaging. In J. C. Masdeu \u0026amp; R. G. Gonz\u0026aacute;lez (Eds.), \u003cem\u003eHandbook of Clinical Neurology\u003c/em\u003e (3rd ed., Vol. 135, pp. 61\u0026ndash;92). Elsevier B.V. https://doi.org/10.1016/B978-0-444-53485-9.00004-0\u003c/li\u003e\n\u003cli\u003eDierker, D., Roland, J. L., Kamran, M., Rutlin, J., Hacker, C. D., Marcus, D. S., Milchenko, M., Miller-Thomas, M. M., Benzinger, T. L., Snyder, A. Z., Leuthardt, E. C., \u0026amp; Shimony, J. S. (2017). Resting-state Functional Magnetic Resonance Imaging in Presurgical Functional Mapping: Sensorimotor Localization. \u003cem\u003eNeuroimaging Clinics of North America\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(4), 621\u0026ndash;633. https://doi.org/10.1016/J.NIC.2017.06.011\u003c/li\u003e\n\u003cli\u003eIraji, A., Fu, Z., Faghiri, A., Duda, M., Chen, J., Rachakonda, S., DeRamus, T., Kochunov, P., Adhikari, B. M., Belger, A., Ford, J. M., Mathalon, D. H., Pearlson, G. D., Potkin, S. G., Preda, A., Turner, J. A., Erp, T. G. M. van, Bustillo, J. R., Yang, K., \u0026hellip; Calhoun, V. D. (2022). Canonical and Replicable Multi-Scale Intrinsic Connectivity Networks in 100k+ Resting-State fMRI Datasets. \u003cem\u003eBioRxiv\u003c/em\u003e, 2022.09.03.506487. https://doi.org/10.1101/2022.09.03.506487\u003c/li\u003e\n\u003cli\u003eJurkiewicz, M. T., Crawley, A. P., \u0026amp; Mikulis, D. J. (2018). Is rest really rest? resting-state functional connectivity during rest and motor task paradigms. \u003cem\u003eBrain Connectivity\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(5), 268\u0026ndash;275. https://doi.org/10.1089/brain.2017.0495\u003c/li\u003e\n\u003cli\u003eLakhani, D. A., Sabsevitz, D. S., Chaichana, K. L., Qui\u0026ntilde;ones-Hinojosa, A., \u0026amp; Middlebrooks, E. H. (2023). Current State of Functional MRI in the Presurgical Planning of Brain Tumors. \u003cem\u003eRadiology: Imaging Cancer\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(6). https://doi.org/10.1148/RYCAN.230078/ASSET/IMAGES/LARGE/RYCAN.230078.FIG5.JPEG\u003c/li\u003e\n\u003cli\u003eLeh\u0026eacute;ricy, S., Duffau, H., Cornu, P., Capelle, L., Pidoux, B., Carpentier, A., Auliac, S., Clemenceau, S., Sichez, J. P., Bitar, A., Valery, C. A., Van Effenterre, R., Faillot, T., Srour, A., Fohanno, D., Philippon, J., Le Bihan, D., \u0026amp; Marsault, C. (2000). Correspondence between functional magnetic resonance imaging somatotopy and individual brain anatomy of the central region: comparison with intraoperative stimulation in patients with brain tumors. \u003cem\u003eJournal of Neurosurgery\u003c/em\u003e, \u003cem\u003e92\u003c/em\u003e(4), 589\u0026ndash;598. https://doi.org/10.3171/JNS.2000.92.4.0589\u003c/li\u003e\n\u003cli\u003eLin, Q.-H., Liu, J., Zheng, Y.-R., Liang, H., \u0026amp; Calhoun, V. D. (2010). Semiblind Spatial ICA of fMRI Using Spatial Constraints. \u003cem\u003eHuman Brain Mapping\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e, 1076\u0026ndash;1088. https://doi.org/10.1002/hbm.20919\u003c/li\u003e\n\u003cli\u003eLu, J., Zhang, H., Hameed, N. U. F., Zhang, J., Yuan, S., Qiu, T., Shen, D., Wu, J., \u0026amp; Lu, K. J. (2017). An automated method for identifying an independent component analysis-based language-related resting-state network in brain tumor subjects for surgical planning. \u003cem\u003eScientific Reports\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(13769). https://doi.org/10.1038/s41598-017-14248-5\u003c/li\u003e\n\u003cli\u003eLu, W., \u0026amp; Rajapakse, J. C. (2005). Approach and applications of constrained ICA. \u003cem\u003eIEEE Transactions on Neural Networks\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(1), 203\u0026ndash;212. https://doi.org/10.1109/TNN.2004.836795\u003c/li\u003e\n\u003cli\u003eNitschke, M. F., Melchert, U. H., Hahn, C., Otto, V., Arnold, H., Herrmann, H. D., Nowak, G., Westphal, M., \u0026amp; Wessel, K. (1998). Preoperative functional magnetic resonance imaging (fMRI) of the motor system in patients with tumours in the parietal lobe. \u003cem\u003eActa Neurochirurgica\u003c/em\u003e, \u003cem\u003e140\u003c/em\u003e(12), 1223\u0026ndash;1229. https://doi.org/10.1007/S007010050242\u003c/li\u003e\n\u003cli\u003eOtt, C., Rosengarth, K., Doenitz, C., Hoehne, J., Wendl, C., Dodoo-Schittko, F., Lang, E., Schmidt, N. O., \u0026amp; Goldhacker, M. (2021). Preoperative assessment of language dominance through combined resting-state and task-based functional magnetic resonance imaging. \u003cem\u003eJournal of Personalized Medicine\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(12), 1\u0026ndash;14. https://doi.org/10.3390/jpm11121342\u003c/li\u003e\n\u003cli\u003eParker Jones, O., Voets, N. L., Adcock, J. E., Stacey, R., \u0026amp; Jbabdi, S. (2017). Resting connectivity predicts task activation in pre-surgical populations. \u003cem\u003eNeuroImage: Clinical\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e, 378\u0026ndash;385. https://doi.org/10.1016/j.nicl.2016.12.028\u003c/li\u003e\n\u003cli\u003ePujol, J., Conesa, G., Deus, J., L\u0026oacute;pez-Obarrio, L., Isamat, F., \u0026amp; Capdevila, A. (1998). Clinical application of functional magnetic resonance imaging in presurgical identification of the central sulcus. \u003cem\u003eJournal of Neurosurgery\u003c/em\u003e, \u003cem\u003e88\u003c/em\u003e(5), 863\u0026ndash;869. https://doi.org/10.3171/JNS.1998.88.5.0863\u003c/li\u003e\n\u003cli\u003eQiu, T. M., Gong, F. Y., Gong, X., Wu, J. S., Lin, C. P., Biswal, B. B., Zhuang, D. X., Yao, C. J., Zhang, X. L., Lu, J. F., Zhu, F. P., Mao, Y., \u0026amp; Zhou, L. F. (2017). Real-Time Motor Cortex Mapping for the Safe Resection of Glioma: An Intraoperative Resting-State fMRI Study. \u003cem\u003eAJNR. American Journal of Neuroradiology\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(11), 2146\u0026ndash;2152. https://doi.org/10.3174/AJNR.A5369\u003c/li\u003e\n\u003cli\u003eQiu, T. ming, Yan, C. gan, Tang, W. jun, Wu, J. song, Zhuang, D. xiao, Yao, C. jun, Lu, J. feng, Zhu, F. ping, Mao, Y., \u0026amp; Zhou, L. fu. (2014). Localizing hand motor area using resting-state fMRI: validated with direct cortical stimulation. \u003cem\u003eActa Neurochirurgica\u003c/em\u003e, \u003cem\u003e156\u003c/em\u003e(12), 2295\u0026ndash;2302. https://doi.org/10.1007/S00701-014-2236-0/FIGURES/2\u003c/li\u003e\n\u003cli\u003eRolinski, R., You, X., Gonzalez-Castillo, J., Norato, G., Reynolds, R. C., Inati, S. K., \u0026amp; Theodore, W. H. (2020). Language lateralization from task-based and resting state functional MRI in patients with epilepsy. \u003cem\u003eHuman Brain Mapping\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(11), 3133\u0026ndash;3146. https://doi.org/10.1002/hbm.25003\u003c/li\u003e\n\u003cli\u003eRorden, C., \u0026amp; Brett, M. (2000). Stereotaxic display of brain lesions. \u003cem\u003eBehavioural Neurology\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e, 191\u0026ndash;200. https://doi.org/10.1155/2000/421719\u003c/li\u003e\n\u003cli\u003eRosazza, C., Aquino, D., \u0026rsquo;incerti, D., Cordella, L., \u0026amp; Andronache, R. (2014). Preoperative Mapping of the Sensorimotor Cortex: Comparative Assessment of Task-Based and Resting-State fMRI. \u003cem\u003ePLoS ONE\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(6), 98860. https://doi.org/10.1371/journal.pone.0098860\u003c/li\u003e\n\u003cli\u003eSair, H. I., Agarwal, S., \u0026amp; Pillai, J. J. (2017). Application of Resting State Functional MR Imaging to Presurgical Mapping: Language Mapping. \u003cem\u003eNeuroimaging Clinics of North America\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(4), 635\u0026ndash;644. https://doi.org/10.1016/j.nic.2017.06.003\u003c/li\u003e\n\u003cli\u003eSanju\u0026aacute;n, A., Bustamante, J. C., Forn, C., Ventura-Campos, N., Barr\u0026oacute;s-Loscertales, A., Mart\u0026iacute;nez, J. C., Villanueva, V., \u0026amp; \u0026Aacute;vila, C. (2010). Comparison of two fMRI tasks for the evaluation of the expressive language function. \u003cem\u003eNeuroradiology\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(5), 407\u0026ndash;415. https://doi.org/10.1007/s00234-010-0667-8\u003c/li\u003e\n\u003cli\u003eSeitzman, B. A., Snyder, A. Z., Leuthardt, E. C., \u0026amp; Shimony, J. S. (2019). The State of Resting State Networks. \u003cem\u003eTopics in Magnetic Resonance Imaging : TMRI\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(4), 189. https://doi.org/10.1097/RMR.0000000000000214\u003c/li\u003e\n\u003cli\u003eSparacia, G., Parla, G., Cannella, R., Perri, A., Lo Re, V., Mamone, G., Miraglia, R., Torregrossa, F., \u0026amp; Grasso, G. (2019). Resting-State Functional Magnetic Resonance Imaging for Brain Tumor Surgical Planning: Feasibility in Clinical Setting. \u003cem\u003eWorld Neurosurgery\u003c/em\u003e, \u003cem\u003e131\u003c/em\u003e, 356\u0026ndash;363. https://doi.org/10.1016/j.wneu.2019.07.022\u003c/li\u003e\n\u003cli\u003eSparacia, G., Parla, G., Mamone, G., Caruso, M., Torregrossa, F., \u0026amp; Grasso, G. (2021). Resting-State Functional Magnetic Resonance Imaging for Surgical Neuro-Oncology Planning: Towards a Standardization in Clinical Settings. \u003cem\u003eBrain Sciences 2021, Vol. 11, Page 1613\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(12), 1613. https://doi.org/10.3390/BRAINSCI11121613\u003c/li\u003e\n\u003cli\u003eTie, Y., Rigolo, L., Norton, I. H., Huang, R. Y., Wu, W., Orringer, D., Mukundan, S., \u0026amp; Golby, A. J. (2014). Defining Language Networks From Resting-State fMRI for Surgical Planning-A Feasibility Study. \u003cem\u003eHuman Brain Mapping\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e, 1018\u0026ndash;1030. https://doi.org/10.1002/hbm.22231\u003c/li\u003e\n\u003cli\u003eTyndall, A. J., Reinhardt, J., Tronnier, V., Mariani, L., \u0026amp; Stippich, C. (2017). Presurgical motor, somatosensory and language fMRI: Technical feasibility and limitations in 491 patients over 13 years. \u003cem\u003eEuropean Radiology\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(1), 267\u0026ndash;278. https://doi.org/10.1007/S00330-016-4369-4/TABLES/3\u003c/li\u003e\n\u003cli\u003eVillar‐Rodr\u0026iacute;guez, E., Palomar‐Garc\u0026iacute;a, M., Hern\u0026aacute;ndez, M., Adri\u0026aacute;n‐Ventura, J., Olcina‐Sempere, G., Parcet, M., \u0026amp; \u0026Aacute;vila, C. (2020). Left‐handed musicians show a higher probability of atypical cerebral dominance for language. \u003cem\u003eHuman Brain Mapping\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(8), 2048. https://doi.org/10.1002/HBM.24929\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"brain-imaging-and-behavior","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bior","sideBox":"Learn more about [Brain Imaging and Behavior](https://www.springer.com/journal/11682)","snPcode":"11682","submissionUrl":"https://submission.nature.com/new-submission/11682/3","title":"Brain Imaging and Behavior","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"resting-state fMRI, motor cortex, ICA, tumor","lastPublishedDoi":"10.21203/rs.3.rs-8287975/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8287975/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cb\u003eBackground\u003c/b\u003e\u003c/p\u003e \u003cp\u003eAlthough resting-state fMRI is a promising alternative to task-fMRI in presurgical mapping, protocols to localize simultaneously and accurately both language and motor areas are still missing.\u003c/p\u003e\u003cp\u003e\u003cb\u003ePurpose\u003c/b\u003e\u003c/p\u003e \u003cp\u003eTo propose a methodology based on the connectivity analysis to identify motor and language areas using a single 6-minute fMRI language task in presurgical patients with space-occupying lesions.\u003c/p\u003e\u003cp\u003e\u003cb\u003eMethods\u003c/b\u003e\u003c/p\u003e \u003cp\u003eIn a retrospective study and using a language fMRI task (verb generation), we established limits of the motor cortex in 40 presurgical patients. Single-subject spatially-constrained ICA was performed on verb generation scans to extract somatomotor network. Sensitivity, and specificity between predicted SMN and hand motor task were calculated. Variables effect was analyzed through ANOVA.\u003c/p\u003e\u003cp\u003e\u003cb\u003eResults\u003c/b\u003e\u003c/p\u003e \u003cp\u003eForty patients (mean age, 40.50\u0026thinsp;\u0026plusmn;\u0026thinsp;13.99 [standard deviation]; 21 men) diagnosed with a space-occupying lesion were included. Somatomotor network extracted from spatially-constrained ICA and language lateralization from task-fMRI were identified in 40/40 (100%) patients. Using the motor task as reference standard, ipsilesional voxel-to-voxel mean sensitivity, and specificity for spatially-constrainted ICA at voxel level was 78%, and 75%, respectively. No significant differences were found between ipsilesional and contralesional hemispheres in sensitivity or specificity. Right contralesional hemisphere, higher scanner field strength, and typical language lateralization, showed higher sensitivity values.\u003c/p\u003e\u003cp\u003e\u003cb\u003eConclusions\u003c/b\u003e\u003c/p\u003e \u003cp\u003eOur study demonstrates the possibility of identifying somatomotor network connectivity from a language task while determining language lateralization in patients with space-occupying lesions. Our technique is a promising alternative to reduce scanning time and cost as only one fMRI sequence is needed to determine the two main eloquent functions.\u003c/p\u003e","manuscriptTitle":"Spatially constrained ICA from a language task predicts Sensorimotor Network in patients with tumors","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-22 17:18:54","doi":"10.21203/rs.3.rs-8287975/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-08T15:35:32+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-11T09:21:40+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"339789301997456178517200321194968666572","date":"2026-03-29T23:33:27+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-18T13:33:32+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-12-16T05:18:06+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-12-09T08:52:46+00:00","index":"","fulltext":""},{"type":"submitted","content":"Brain Imaging and Behavior","date":"2025-12-05T12:45:06+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"brain-imaging-and-behavior","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bior","sideBox":"Learn more about [Brain Imaging and Behavior](https://www.springer.com/journal/11682)","snPcode":"11682","submissionUrl":"https://submission.nature.com/new-submission/11682/3","title":"Brain Imaging and Behavior","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"18ec3f74-8232-471c-b27b-3d865a52e57a","owner":[],"postedDate":"December 22nd, 2025","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-08T15:35:32+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-21T21:53:20+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-22 17:18:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8287975","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8287975","identity":"rs-8287975","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}