A deep representation learning model to predict response to vagus nerve stimulation

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.

Authors

Institutions

Publication Details

Journal
Nature Communications
Published
2026-04-07
DOI
https://doi.org/10.1038/s41467-026-71555-0
Citations
1
Primary Topic
Vagus Nerve Stimulation Research
Type
article
Field-Weighted Citation Impact
16.08
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

A deep representation learning model to predict response to vagus nerve stimulation

Carolina Gorodetsky, Madison Remick, Sean M. Lew, Robert J. Bollo et al.
1 citations
Nature Communications
Vagus Nerve Stimulation Research
16.08
article

A deep representation learning model to predict response to vagus nerve stimulation

Carolina Gorodetsky, Madison Remick, Sean M. Lew, Robert J. Bollo, Toba N. Niazi, Sami Obaïd, Dewi Schrader, Timur H. Latypov, Jeffrey S. Raskin, Alexandre Berger, Melissa A. LoPresti, Karim Mithani, Logi Vidarsson, Nebras M. Warsi, Nisha Gadgil, Sebastian C. Coleman, Howard L. Weiner, Gregory W. Albert, Guillaume Theaud, Emefa Akwayena, Shelly Wang, Hrishikesh Suresh, Alexander G. Weil, George M. Ibrahim, Simeon M. Wong, Farbod Niazi, Vicki Li, Flavia Venetucci Gouveia, Lauren Erdman, Philippe Major, Margot J. Taylor, Jurgen Germann, Elysa Widjaja, Aria Fallah, James T. Rutka, Vann Chau, Diana Aum, Matthew D. Smyth, Elizabeth Donner, Aristides Hadjinicolaou, Alexandre Boutet, Birgit Ertl-Wagner, Shelly Weiss, Taylor J. Abel, Puneet Jain, Jaeyoung Kang
article en
1 citations

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.

Nature Communications
Cincinnati Children's Hospital Medical Center (US), Arkansas Children's Hospital (US), St. Louis Children's Hospital (US), University Health Network (CA), Johns Hopkins University (US), University of Toronto (CA), Medical College of Wisconsin (US), University of Utah (US), Hospital for Sick Children (CA), Centre Hospitalier Universitaire Sainte-Justine (CA), Riley Hospital for Children (US), University of Rochester Medical Center (US), Johns Hopkins All Children's Hospital (US), Children's Hospital of Pittsburgh (US), Mattel Children's Hospital (US), Comer Children's Hospital (US), SickKids Foundation (CA), Centre Hospitalier de l’Université de Montréal (CA), Miami Children's Hospital (US), Texas Children's Hospital (US), BC Children's Hospital (CA), Ontario Brain Institute (CA), Krembil Brain Institute, Université de Montréal (CA)
Peace, Justice and strong institutions
Openalex Percentile: Top 1%
Vagus Nerve Stimulation Research
16.08
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.