EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Multimodal Machine Learning to Personalize Transcutaneous Spinal Cord Stimulation for Stroke Rehabilitation

Multimodales Machine Learning zur Personalisierung der transkutanen spinalen Rückenmarksstimulation für die Schlaganfallrehabilitation

Ameen Kishta, Rushmin Khazanchi, Nicole Veit, Mitchell Tillman, Kelly McKenzie, Shreya Aalla, Jasmine Glenn-Hunt, Chen Yang, Luigi Privitera, Xin S Yu, Jai Chundi, Francesco Lanotte und 3 weitere

Digital Biomarkers · 05.05.2026 · 10 (1) · 151-164

DOI: 10.1159/000552236 ↗

Auf einen Blick

Bei 20 Erwachsenen mit chronischem Schlaganfall sagte ein überwacht trainiertes ML-Modell die optimalen tSCS-Parameter für Frequenz und Intensität mit guter Trennschärfe voraus.

Fragestellung

Kann ein supervised-Machine-Learning-Ansatz die optimalen tSCS-Frequenz- und Intensitätsparameter bei chronischem Schlaganfall vorhersagen?

Methode

20 Erwachsene mit chronischem hemiparetischem Schlaganfall; Baseline plus fünf Interventionssitzungen mit zufällig variierter Frequenz und Intensität; nested leave-one-subject-out cross-validation; AUROC-Berechnung.

Ergebnisse

AUROC 0,86 [0,75–0,94] für die Frequenzvorhersage und 0,82 [0,69–0,94] für die Intensitätsvorhersage. Die wichtigsten Merkmale umfassten spinal motor evoked potentials, Wearablesensoren und Demografie.

Schlussfolgerung der Autor:innen

Die Autoren schließen, dass supervised learning individuelle tSCS-Parameter aus demografischen Daten und Basissensoren vorhersagen kann und damit ein Schritt zur datengetriebenen Personalisierung der Neuromodulation ist.

Kernaussagen

  • 20 Erwachsene mit chronischem Schlaganfall
  • AUROC 0,86 für Frequenz, 0,82 für Intensität
  • Multimodale Merkmale trugen zur Vorhersage bei

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift Digital Biomarkers, online erschienen am 05.05.2026. Verfasst von Ameen Kishta und 14 weiteren Autor:innen. Beteiligte Einrichtung: Max Näder Lab for Rehabilitation Technologies and Outcomes Research, Shirley Ryan AbilityLab, Chicago, IL, USA. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: NICHD NIH HHS.

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Several studies have identified transcutaneous spinal cord stimulation (tSCS) as a noninvasive neuromodulation technique for improving motor function in individuals with neurological disorders, including stroke. Despite a plethora of preliminary findings, there remains no standardized protocol regarding the optimal tSCS parameters tailored to patients with stroke. The objective of this study was to employ a supervised machine learning (ML) approach to determine the optimal tSCS frequency and intensity parameters for patients with chronic stroke by leveraging data from single-day interventions that systematically varied in frequency and intensity across four stimulation conditions.

Methods

Twenty adults with chronic hemiparetic stroke (mean age 53.3 ± 10.8 years; 13 males, 7 females) who were ≥6 months post-stroke was enrolled, excluding individuals with multiple strokes, severe spasticity, or implanted devices. Each participant participated in a baseline session followed by five intervention sessions, during which the frequency and intensity of stimulation were varied randomly. Multimodal sensors, including surface EMG, inertial measurement unit-based kinematics, and spatiotemporal gait parameters, were used to quantify acute changes in gait symmetry, and optimal stimulation frequency and intensity were defined as those yielding the greatest improvement in the combined gait asymmetry metric. A supervised machine-learning classifier was trained using nested leave-one-subject-out cross-validation to predict the optimal stimulation frequency and intensity to maximize differences from the baseline data alone. AUROC values were calculated for the frequency and intensity predictions.

Results

The ML models achieved AUROC values of 0.86 [0.75–0.94] for frequency prediction and 0.82 [0.69–0.94] for intensity prediction. The top ten predictive features for each model spanned with spinal motor evoked potentials, wearable sensors, and demographic domains, highlighting multimodal contributions to stimulation optimization.

Conclusion

These findings demonstrate that supervised learning can predict individualized tSCS parameters from demographic data and baseline sensor features that yield the greatest improvement in gait symmetry after stroke, representing a promising step toward the data-driven personalization of neuromodulation therapy in neurorehabilitation.

Schlagwörter

Stichwörter der Autor:innen

Digital biomarkersElectromyographyGait analysisInertial measurement unitsMultimodal machine learningPersonalized neuromodulationStroke rehabilitationTranscutaneous spinal cord stimulationWearable sensors

Autor:innen und Institutionen

  • Max Näder Lab for Rehabilitation Technologies and Outcomes Research, Shirley Ryan AbilityLab, Chicago, IL, USA.
  • Feinberg School of Medicine, Northwestern University, Chicago, IL, USA.
  • Biomedical Engineering, McCormick School of Engineering, Northwestern University, Evanston, IL, USA.
  • Center for Sustainable Future Technologies, Istituto Italiano di Tecnologia, Turin, Italy.
  • Physical Medicine and Rehabilitation, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA.
  • Hines VA Medical Center, Maywood, IL, USA.

Förderung

  • NICHD NIH HHS

Zitieren

APA

Kishta, A., Khazanchi, R., Veit, N., Tillman, M., McKenzie, K., Aalla, S., Glenn-Hunt, J., Yang, C., Privitera, L., Yu, X. S., Chundi, J., Lanotte, F., Lieber, R. L., Roth, E., & Jayaraman, A. (2026). Multimodal Machine Learning to Personalize Transcutaneous Spinal Cord Stimulation for Stroke Rehabilitation. Digital Biomarkers, 10(1), 151-164. https://doi.org/10.1159/000552236

Vancouver

Kishta A, Khazanchi R, Veit N, Tillman M, McKenzie K, Aalla S, et al. Multimodal Machine Learning to Personalize Transcutaneous Spinal Cord Stimulation for Stroke Rehabilitation. Digital Biomarkers. 2026;10(1):151-164. doi:10.1159/000552236

BibTeX
@article{Kishta2026,
  author = {Kishta, Ameen and Khazanchi, Rushmin and Veit, Nicole and Tillman, Mitchell and McKenzie, Kelly and Aalla, Shreya and Glenn-Hunt, Jasmine and Yang, Chen and Privitera, Luigi and Yu, Xin S and Chundi, Jai and Lanotte, Francesco and Lieber, Richard L and Roth, Elliot and Jayaraman, Arun},
  title = {Multimodal Machine Learning to Personalize Transcutaneous Spinal Cord Stimulation for Stroke Rehabilitation},
  journal = {Digital Biomarkers},
  year = {2026},
  volume = {10},
  number = {1},
  pages = {151--164},
  doi = {10.1159/000552236}
}
Kostenlos starten →