Abstract
Implantable neurotechnologies are increasingly used to reduce seizure burden in pediatric epilepsy. Vagus nerve stimulation (VNS), the most common option, is effective for only half of patients, with no means to predict outcome prior to surgery. As a result, many children undergo invasive and costly procedures without benefit. Although T1-weighted magnetic resonance imaging (T1w) is routinely acquired presurgically and may capture structural brain differences relevant to treatment outcome, its high dimensionality relative to sample sizes has limited its utility in predictive modelling. To address this challenge, we present VQ-VNS, a deep representation learning model to predict VNS outcome based on preoperative T1w (n = 263). First, we present data from the largest paediatric VNS cohort (n = 1046), wherein presurgical clinical data could not predict response (AUC 0.54,p > 0.99). Next, VQ-VNS was pretrained on 7433 T1w images to learn compact anatomical representations enabling its classifier to predict VNS response (AUC = 0.73,p = 0.007). Model predictions localized to serotonin-rich brain regions and inferred large-scale disruptions in network connectivity among non-responders. This biologically interpretable predictor based on routine structural imaging improves upon current clinical decision-making.
Data availability
Due to ethical and legal concerns for patient privacy, individual participant data cannot be made public. Deidentified data used in this study can be made available subject to the policies and procedures of the institution from which the data were collected. Data requests should be sent to the corresponding author. Requests for data will be reviewed within 4 weeks.
Code availability
All code for the machine learning model has been made publicly available on GitHub (https://github.com/gmilab/VQVNS102) and HuggingFace (https://huggingface.co/hsuresh/vqvns).
References
Aaberg, K. M. et al. Incidence and prevalence of childhood epilepsy: A nationwide cohort study. Pediatrics 139, e20163908 (2017).
Ryvlin, P. et al. Long-term surveillance of SUDEP in drug-resistant epilepsy patients treated with VNS therapy. Epilepsia 59, 562–572 (2018).
Apantaku, G. O. et al. Clinician preferences for neurotechnologies in pediatric drug-resistant epilepsy: A discrete choice experiment. Epilepsia 63, 2338–2349 (2022).
Jain, P. & Arya, R. Vagus nerve stimulation and seizure outcomes in pediatric refractory epilepsy: Systematic review and meta-analysis. Neurology 96, 1041–1051 (2021).
Soleman, J. et al. Improved quality of life and cognition after early vagal nerve stimulator implantation in children. Epilepsy Behav. 88, 139–145 (2018).
Muthiah, N. et al. Development of a clinical model to predict vagus nerve stimulation response in pediatric patients with drug-resistant epilepsy. J. Neurosurg. Pediatr. 31, 476–483 (2023).
Kim, M.-J. et al. An interictal EEG can predict the outcome of vagus nerve stimulation therapy for children with intractable epilepsy. Childs Nerv. Syst. 33, 145–151 (2017).
Clifford, H. J. et al. Vagus nerve stimulation for epilepsy: A narrative review of factors predictive of response. Epilepsia 65, 3441–3456 (2024).
Reid, S. A. Surgical technique for implantation of the neurocybernetic prosthesis. Epilepsia 31, S38–S39 (1990).
Bernasconi, A. et al. Recommendations for the use of structural magnetic resonance imaging in the care of patients with epilepsy: A consensus report from the International League Against Epilepsy Neuroimaging Task Force. Epilepsia 60, 1054–1068 (2019).
Berger, A. et al. Identifying responders to vagus nerve stimulation based on microstructural features of thalamocortical tracts in drug-resistant epilepsy. Neurotherapeutics 21, e00422 (2024).
Ibrahim, G. M. et al. Presurgical thalamocortical connectivity is associated with response to vagus nerve stimulation in children with intractable epilepsy. NeuroImage Clin. 16, 634–642 (2017).
Mithani, K. et al. Connectomic profiling identifies responders to vagus nerve stimulation. Ann. Neurol. 86, 743–753 (2019).
Arya, R. et al. Predictors of response to vagus nerve stimulation in childhood-onset medically refractory epilepsy. J. Child Neurol. 29, 1652–1659 (2014).
Ghaemi, K. et al. Vagus nerve stimulation: outcome and predictors of seizure freedom in long-term follow-up. Seizure 19, 264–268 (2010).
Guo, M. et al. Vagus nerve stimulation for pharmacoresistant epilepsy secondary to encephalomalacia: A single-center retrospective study. Front. Neurol. 13, 1074997 (2022).
Xie, H. et al. Efficacy of vagus nerve stimulation in 95 children of drug-resistant epilepsy with structural etiology. Epilepsy Behav. 140, 109107 (2023).
Guo, Z. et al. Brain-clinical signatures for vagus nerve stimulation response. CNS Neurosci. Ther. 29, 855–865 (2023).
van den Oord, A., Vinyals, O. & Kavukcuoglu, K. Neural discrete representation learning. arXiv [cs.LG] (2017) https://doi.org/10.48550/ARXIV.1711.00937.
Rombach, R., Blattmann, A., Lorenz, D., Esser, P. & Ommer, B. High-resolution image synthesis with latent diffusion models. arXiv [cs.CV] (2021).
Tudosiu, P.-D. et al. Realistic morphology-preserving generative modelling of the brain. Nat. Mach. Intell. 6, 811–819 (2024).
Pinaya, W. H. L. et al. Brain imaging generation with Latent Diffusion Models. arXiv [eess.IV] (2022) https://doi.org/10.48550/ARXIV.2209.07162.
Huang, S.-C. et al. Self-supervised learning for medical image classification: a systematic review and implementation guidelines. NPJ Digit. Med. 6, 74 (2023).
Miller, K. L. et al. Multimodal population brain imaging in the UK Biobank prospective epidemiological study. Nat. Neurosci. 19, 1523–1536 (2016).
Mills, K. L. et al. Structural brain development between childhood and adulthood: Convergence across four longitudinal samples. Neuroimage 141, 273–281 (2016).
Jia, X., Ren, L. & Cai, J. Clinical implementation of AI technologies will require interpretable AI models. Med. Phys. 47, 1–4 (2020).
van der Velden, B. H. M., Kuijf, H. J., Gilhuijs, K. G. A. & Viergever, M. A. Explainable artificial intelligence (XAI) in deep learning-based medical image analysis. Med. Image Anal. 79, 102470 (2022).
Cortes, C. & Vapnik, V. Support-vector networks. Mach. Learn. 20, 273–297 (1995).
Selvaraju, R. R. et al. Grad-CAM: Visual explanations from deep networks via gradient-based localization. Int. J. Comput. Vis. 128, 336–359 (2020).
Rodríguez-Pérez, R. & Bajorath, J. Interpretation of machine learning models using shapley values: application to compound potency and multi-target activity predictions. J. Comput. Aided Mol. Des. 34, 1013–1026 (2020).
McInnes, L., Healy, J. & Melville, J. UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction. arXiv [stat.ML] (2018).
Rutecki, P. Anatomical, physiological, and theoretical basis for the antiepileptic effect of vagus nerve stimulation. Epilepsia 31, S1–S6 (1990).
Zhang, R., Isola, P., Efros, A. A., Shechtman, E. & Wang, O. The unreasonable effectiveness of deep features as a perceptual metric. in 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (IEEE, 2018). https://doi.org/10.1109/cvpr.2018.00068.
Dhariwal, P. et al. Jukebox: A generative model for music. arXiv [eess.AS] (2020).
Wang, T.-C. et al. High-resolution image synthesis and semantic manipulation with conditional GANs. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (IEEE, 2018). https://doi.org/10.1109/cvpr.2018.00917.
Wang, Z., Simoncelli, E. P. & Bovik, A. C. Multiscale structural similarity for image quality assessment. In The 37th Asilomar Conference on Signals, Systems & Computers, 2003 (IEEE, 2004). https://doi.org/10.1109/acssc.2003.1292216.
Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B. & Hochreiter, S. GANs trained by a two time-scale update rule converge to a local Nash equilibrium. arXiv [cs.LG] (2017).
Billot, B. et al. SynthSeg: Segmentation of brain MRI scans of any contrast and resolution without retraining. Med. Image Anal. 86, 102789 (2023).
Arski, O. N. et al. Connectomic profiles and cognitive trajectories after epilepsy surgery in children. Neurology 98, e2233–e2244 (2022).
Hansen, J. Y. et al. Mapping neurotransmitter systems to the structural and functional organization of the human neocortex. Nat. Neurosci. 25, 1569–1581 (2022).
Lozano, A. M. & Lipsman, N. Probing and regulating dysfunctional circuits using deep brain stimulation. Neuron 77, 406–424 (2013).
Fox, M. D. Mapping symptoms to brain networks with the human connectome. N. Engl. J. Med. 379, 2237–2245 (2018).
Narayanan, J. T. et al. Cerebral activation during vagus nerve stimulation: a functional MR study. Epilepsia 43, 1509–1514 (2002).
Suk, H.-I., Lee, S.-W. & Shen, D. & Alzheimer’s Disease Neuroimaging Initiative. Hierarchical feature representation and multimodal fusion with deep learning for AD/MCI diagnosis. Neuroimage 101, 569–582 (2014).
Davatzikos, C. Why voxel-based morphometric analysis should be used with great caution when characterizing group differences. Neuroimage 23, 17–20 (2004).
Lecun, Y., Bottou, L., Bengio, Y. & Haffner, P. Gradient-based learning applied to document recognition. Proc. IEEE Inst. Electr. Electron. Eng. 86, 2278–2324 (1998).
Henry, T. R. et al. Acute blood flow changes and efficacy of vagus nerve stimulation in partial epilepsy. Neurology 52, 1166–1173 (1999).
Ito, S.-I. Visceral region in the rat primary somatosensory cortex identified by vagal evoked potential. J. Comp. Neurol. 444, 10–24 (2002).
Ko, D. et al. Vagus nerve stimulation activates central nervous system structures in epileptic patients during PET H2(15)O blood flow imaging. Neurosurgery 39, 426–421 (1996).
Mithani, K. et al. Somatosensory evoked fields predict response to vagus nerve stimulation. NeuroImage Clin. 26, 102205 (2020).
Hulsey, D. R., Shedd, C. M., Sarker, S. F., Kilgard, M. P. & Hays, S. A. Norepinephrine and serotonin are required for vagus nerve stimulation directed cortical plasticity. Exp. Neurol. 320, 112975 (2019).
Krahl, S. E. & Clark, K. B. Vagus nerve stimulation for epilepsy: A review of central mechanisms. Surg. Neurol. Int. 3, S255–S259 (2012).
Aaronson, S. T. et al. A 5-year observational study of patients with treatment-resistant depression treated with vagus nerve stimulation or treatment as usual: Comparison of response, remission, and suicidality. Am. J. Psychiatry 174, 640–648 (2017).
Furmaga, H., Shah, A. & Frazer, A. Serotonergic and noradrenergic pathways are required for the anxiolytic-like and antidepressant-like behavioral effects of repeated vagal nerve stimulation in rats. Biol. Psychiatry 70, 937–945 (2011).
Fahoum, F. et al. VNS parameters for clinical response in Epilepsy. Brain Stimul. 15, 814–821 (2022).
Mithani, K. et al. Long-term outcomes of vagus nerve stimulation in children with drug resistant epilepsy are driven by early benefit. Brain Stimul. 18, 246–248 (2025).
Berger, A. et al. How Is the Norepinephrine System Involved in the Antiepileptic Effects of Vagus Nerve Stimulation?. Front. Neurosci. 15, 790943 (2021).
De Taeye, L. et al. The P3 event-related potential is a biomarker for the efficacy of vagus nerve stimulation in patients with epilepsy. Neurotherapeutics 11, 612–622 (2014).
Siegel, L. et al. Connectomic profiling and Vagus nerve stimulation Outcomes Study (CONNECTiVOS): a prospective observational protocol to identify biomarkers of seizure response in children and youth. BMJ Open 12, e055886 (2022).
Goodyear, M. D. E., Krleza-Jeric, K. & Lemmens, T. The declaration of Helsinki. BMJ 335, 624–625 (2007).
Cameron, C. et al. The Neuro Bureau Preprocessing Initiative: open sharing of preprocessed neuroimaging data and derivatives. Front. Neuroinform. 7, (2013).
Richie-Halford, A. et al. An analysis-ready and quality controlled resource for pediatric brain white-matter research. Sci. Data 9, 616 (2022).
Reynolds, J. E., Long, X., Paniukov, D., Bagshawe, M. & Lebel, C. Calgary Preschool magnetic resonance imaging (MRI) dataset. Data Brief. 29, 105224 (2020).
Taylor, J. R. et al. The Cambridge Centre for Ageing and Neuroscience (Cam-CAN) data repository: Structural and functional MRI, MEG, and cognitive data from a cross-sectional adult lifespan sample. Neuroimage 144, 262–269 (2017).
Snoek, L. et al. The Amsterdam Open MRI Collection, a set of multimodal MRI datasets for individual difference analyses. Sci. Data 8, 85 (2021).
Taylor, P. N. et al. The imaging database for epilepsy and surgery (IDEAS). Openneuro https://doi.org/10.18112/OPENNEURO.DS005602.V1.0.0 (2024).
Schuch, F. et al. An open presurgery MRI dataset of people with epilepsy and focal cortical dysplasia type II. Sci. Data 10, 475 (2023).
Andersson, J. L. R. & Sotiropoulos, S. N. An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. Neuroimage 125, 1063–1078 (2016).
Hoopes, A., Mora, J. S., Dalca, A. V., Fischl, B. & Hoffmann, M. SynthStrip: skull-stripping for any brain image. Neuroimage 260, 119474 (2022).
Tustison, N. J. et al. N4ITK: improved N3 bias correction. IEEE Trans. Med. Imaging 29, 1310–1320 (2010).
Fischl, B. FreeSurfer. Neuroimage 62, 774–781 (2012).
Reinhold, J. C., Dewey, B. E., Carass, A. & Prince, J. L. Evaluating the impact of intensity normalization on MR image synthesis. Proc. SPIE 10949, (2019).
Avants, B. B., Epstein, C. L., Grossman, M. & Gee, J. C. Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain. Med. Image Anal. 12, 26–41 (2008).
Fonov, V. S., Evans, A. C., McKinstry, R. C., Almli, C. R. & Collins, D. L. Unbiased nonlinear average age-appropriate brain templates from birth to adulthood. Neuroimage 47, S102 (2009).
Akman, C. I. et al. Seizure frequency in children with epilepsy: factors influencing accuracy and parental awareness. Seizure 18, 524–529 (2009).
Shahapure, K. R. & Nicholas, C. Cluster quality analysis using silhouette score. in 2020 IEEE 7th International Conference on Data Science and Advanced Analytics (DSAA) 747–748 (IEEE, 2020).
Guyon, I., Weston, J., Barnhill, S. & Vapnik, V. Gene selection for cancer classification using support vector machines. Mach. Learn. 46, 389–422 (2002).
FsAverage - Free Surfer Wiki. https://surfer.nmr.mgh.harvard.edu/fswiki/FsAverage.
Ghosh, S. S. et al. Evaluating the validity of volume-based and surface-based brain image registration for developmental cognitive neuroscience studies in children 4 to 11 years of age. Neuroimage 53, 85–93 (2010).
Krizhevsky, A., Sutskever, I. & Hinton, G. E. ImageNet classification with deep convolutional neural networks. Commun. ACM 60, 84–90 (2017).
Chen, Z., Badrinarayanan, V., Lee, C.-Y. & Rabinovich, A. GradNorm: Gradient normalization for adaptive loss balancing in deep multitask networks. arXiv [cs.CV] (2017).
Tseng, H.-Y., Jiang, L., Liu, C., Yang, M.-H. & Yang, W. Regularizing generative adversarial networks under limited data. arXiv [cs.LG] (2021).
Kingma, D. P. & Ba, J. Adam: A method for stochastic optimization. arXiv [cs.LG] (2014).
He, K., Zhang, X., Ren, S. & Sun, J. Deep residual learning for image recognition. arXiv [cs.CV] (2015).
Lundberg, S. & Lee, S.-I. A unified approach to interpreting model predictions. arXiv [cs.AI] (2017).
Tustison, N. J. et al. The ANTsX ecosystem for quantitative biological and medical imaging. Sci. Rep. 11, 9068 (2021).
Markello, R. D. et al. Neuromaps: Structural and functional interpretation of brain maps. Nat. Methods 19, 1472–1479 (2022).
Alexander-Bloch, A. F. et al. On testing for spatial correspondence between maps of human brain structure and function. Neuroimage 178, 540–551 (2018).
van der Maaten, L. & Hinton, G. Visualizing data using t-SNE. J. Mach. Learn. Res. 9, 2579–2605 (2008).
Karnath, H.-O., Sperber, C. & Rorden, C. Mapping human brain lesions and their functional consequences. Neuroimage 165, 180–189 (2018).
Theaud, G. et al. TractoFlow: A robust, efficient and reproducible diffusion MRI pipeline leveraging Nextflow & Singularity. Neuroimage 218, 116889 (2020).
Girard, G., Whittingstall, K., Deriche, R. & Descoteaux, M. Towards quantitative connectivity analysis: reducing tractography biases. Neuroimage 98, 266–278 (2014).
St-Onge, E., Daducci, A., Girard, G. & Descoteaux, M. Surface-enhanced tractography (SET). Neuroimage 169, 524–539 (2018).
Daducci, A., Dal Palù, A., Lemkaddem, A. & Thiran, J.-P. COMMIT: Convex optimization modeling for microstructure informed tractography. IEEE Trans. Med. Imaging 34, 246–257 (2015).
Di Tommaso, P. et al. Nextflow enables reproducible computational workflows. Nat. Biotechnol. 35, 316–319 (2017).
Kurtzer, G. M., Sochat, V. & Bauer, M. W. Singularity: Scientific containers for mobility of compute. PLoS One 12, e0177459 (2017).
Hollunder, B. et al. Mapping dysfunctional circuits in the frontal cortex using deep brain stimulation. Nat. Neurosci. 27, 573–586 (2024).
Foulon, C. et al. Advanced lesion symptom mapping analyses and implementation as BCBtoolkit. Gigascience 7, (2018).
Cox, R. W. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput. Biomed. Res. 29, 162–173 (1996).
Cohen, A., Soussand, L., McManus, P. & Fox, M. GSP1000 Preprocessed Connectome. Harvard Dataverse https://doi.org/10.7910/DVN/ILXIKS (2020).
Winkler, A. M., Ridgway, G. R., Webster, M. A., Smith, S. M. & Nichols, T. E. Permutation inference for the general linear model. Neuroimage 92, 381–397 (2014).
Suresh, H. Gmilab/VQVNS: V10. (Zenodo, 2026). https://doi.org/10.5281/ZENODO.18510266.
Acknowledgements
This work was funded by a Project Grant from the Canadian Institutes of Health Research (PJT159561). This study was also partially funding by an investigator-initiated grant from LivaNova PLC (VNS Manufacturer in this study) for prospective data from 4 sites for children treated on-label. However, LivaNova PLC did not play any role in study design, data collection, data analysis, data interpretation, or manuscript preparation. H.S. received a doctoral grant from the Canadian Institute of Health Research - Canada Graduate Scholarship. G.M.I received funding from the Abe Bresover Chair in Functional Neurosurgery.
Author information
Authors and Affiliations
Contributions
H.S. and G.M.I. conceived and designed the study. H.S., K.M., V.L., T.H.L., N.M.W., S.M.W., J.K., L.E., J.G., S.C.C., A.B., G.T., S.O., and G.M.I. curated the data and performed formal analysis. H.S., K.M., S.M.W., and G.M.I. wrote the original draft of the manuscript. H.S., K.M., V.L., T.H.L., N.M.W., S.M.W., J.K., J.G., F.V.G., S.C.C., A.B., V.C., S.W., C.G., E.D., A.G.W., J.T., T.J.A., M.R., E.A., D.S., R.J.B., M.D.S., D.A., S.M.L., S.Wa., T.N.N., A.F., J.S.R., H.L.W., N.G., G.W.A., A.H., P.M., F.N., G.T., S.O., E.W., B.E.W., L.V., M.J.T., A.Bo., J.T.R., M.A.L., P.J., and G.M.I. contributed to review and editing of the manuscript. H.S. and G.M.I. acquired funding. H.S., K.M., T.H.L., N.M.W., S.M.W., J.G., F.V.G., S.C.C., A.B., V.C., S.W., C.G., E.D., A.G.W., J.T., T.J.A., M.R., E.A., D.S., R.J.B., M.D.S., D.A., S.M.L., S.Wa., T.N.N., A.F., J.S.R., H.L.W., N.G., G.W.A., A.H., P.M., F.N., G.T., S.O., E.W., B.E.W., L.V., M.J.T., A.Bo., J.T.R., M.A.L., P.J., and G.M.I. supported project administration and/or provided resources. G.M.I. supervised all aspects of the study.
Corresponding author
Ethics declarations
Competing interests
GMI has received consulting and advisory board fees from LivaNova PLC and Medtronic Inc, and serves on the scientific advisory boards of Synergia Inc and the Paediatric Epilepsy Surgery Alliance. The remaining authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks Riëm El Tahry and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Suresh, H., Mithani, K., Li, V. et al. A deep representation learning model to predict response to vagus nerve stimulation. Nat Commun (2026). https://doi.org/10.1038/s41467-026-71555-0
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-026-71555-0
