EUVIMEDEUVIMED entdecken
Open Access

Human-inspired time-series health evaluation with an adaptive multimodal electronic skin

Changhao Xu iD, Hongkai Zheng iD, Wenzheng Heng iD, Ruixiao Liu iD, John MarionSims iD, Jiahong Li iD, Peng Jin iD, Roland Yingjie Tay iD, Jihong Min iD, Guanzhi Wang iD, Yisong Yue iD, Wei Gao iD

Science Advances · 31.07.2026 · 12 (31)

DOI: 10.1126/sciadv.aeg5606 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Science Advances, erschienen am 31.07.2026. Verfasst von Changhao Xu und 11 weiteren Autor:innen. Beteiligte Einrichtung: Andrew and Peggy Cherng Department of Medical Engineering, Division of Engineering and Applied Science, California Institute of Technology, Pasadena,…. Hauptthemen laut MEDLINE: Wearable Electronic Devices. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 78 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIDCD NIH HHS, NHLBI NIH HHS, National Institutes of Health.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Electronic skin powered with artificial intelligence could enable next-generation robotic and medical devices, yet integrating multimodal sensors and analyzing heterogeneous, multifrequency time series remain challenging. Most wearable machine learning architectures are time-invariant and trained for a specific task, limiting transfer across modalities and users. We present a multimodal electronic skin that captures diverse physiological signs with an adaptive learning framework that rapidly generalizes to unseen tasks with minimal labeled data. Our streamlined end-to-end framework uses a spectral variational autoencoder to denoise and compress multifrequency biosignals into a shared, unified second-wise latent space that preserves the spectral-temporal structure, followed by a transformer to capture temporal dependencies to support diverse downstream tasks with data-efficient learning. We demonstrate robust adaptation with 94.7% accuracy in activity recognition and 90.2% precision in fatigue assessment across various users and daily activities regardless of device and user variations, highlighting a scalable route to generalized physiological time-series analytics and human performance assessments.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Wearable Electronic DevicesHumansMachine LearningDigital HealthAdaptive AlgorithmsAutoencoder

Autor:innen und Institutionen

  • Andrew and Peggy Cherng Department of Medical Engineering, Division of Engineering and Applied Science, California Institute of Technology, Pasadena, CA, USA.
  • Department of Computing + Mathematical Sciences, Division of Engineering and Applied Science, California Institute of Technology, Pasadena, CA, USA.

ORCID-Profile: Changhao Xu, Hongkai Zheng, Wenzheng Heng, Ruixiao Liu, John MarionSims, Jiahong Li, Peng Jin, Roland Yingjie Tay, Jihong Min, Guanzhi Wang, Yisong Yue, Wei Gao

Förderung

  • NIDCD NIH HHS
  • NHLBI NIH HHS
  • National Institutes of Health
  • Office of Naval Research
  • Army Research Office
  • Heritage Medical Research Institute
  • PIMCO Data Science Graduate Fellowship

Zitieren

APA

Xu, C., Zheng, H., Heng, W., Liu, R., MarionSims, J., Li, J., Jin, P., Tay, R. Y., Min, J., Wang, G., Yue, Y., & Gao, W. (2026). Human-inspired time-series health evaluation with an adaptive multimodal electronic skin. Science Advances, 12(31). https://doi.org/10.1126/sciadv.aeg5606

Vancouver

Xu C, Zheng H, Heng W, Liu R, MarionSims J, Li J, et al. Human-inspired time-series health evaluation with an adaptive multimodal electronic skin. Science Advances. 2026;12(31). doi:10.1126/sciadv.aeg5606

BibTeX
@article{Xu2026,
  author = {Xu, Changhao and Zheng, Hongkai and Heng, Wenzheng and Liu, Ruixiao and MarionSims, John and Li, Jiahong and Jin, Peng and Tay, Roland Yingjie and Min, Jihong and Wang, Guanzhi and Yue, Yisong and Gao, Wei},
  title = {Human-inspired time-series health evaluation with an adaptive multimodal electronic skin},
  journal = {Science Advances},
  year = {2026},
  volume = {12},
  number = {31},
  doi = {10.1126/sciadv.aeg5606}
}
Kostenlos starten →