EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis

Imanol Isasa, Mikel Hernandez, Gorka Epelde, Francisco Londoño, Andoni Beristain, Xabat Larrea, Ane Alberdi, Panagiotis Bamidis, Evdokimos Konstantinidis

BMC Medical Informatics and Decision Making · 2024

DOI: 10.1186/s12911-024-02427-0 ↗

Kurz gefasst

Fachartikel in der Zeitschrift BMC Medical Informatics and Decision Making, 2024. Verfasst von Imanol Isasa und 8 weiteren Autor:innen. Beteiligte Einrichtung: Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA), Donostia-San Sebastián, Spain. Hauptthemen laut MEDLINE: Metadata, Privacy. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Horizon 2020 Framework Programme, Department of Education, Universities and Research of the Basque Country.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Synthetic data is an emerging approach for addressing legal and regulatory concerns in biomedical research that deals with personal and clinical data, whether as a single tool or through its combination with other privacy enhancing technologies. Generating uncompromised synthetic data could significantly benefit external researchers performing secondary analyses by providing unlimited access to information while fulfilling pertinent regulations. However, the original data to be synthesized (e.g., data acquired in Living Labs) may consist of subjects’ metadata (static) and a longitudinal component (set of time-dependent measurements), making it challenging to produce coherent synthetic counterparts.

Methods

Three synthetic time series generation approaches were defined and compared in this work: only generating the metadata and coupling it with the real time series from the original data (A1), generating both metadata and time series separately to join them afterwards (A2), and jointly generating both metadata and time series (A3). The comparative assessment of the three approaches was carried out using two different synthetic data generation models: the Wasserstein GAN with Gradient Penalty (WGAN-GP) and the DöppelGANger (DGAN). The experiments were performed with three different healthcare-related longitudinal datasets: Treadmill Maximal Effort Test (TMET) measurements from the University of Malaga (1), a hypotension subset derived from the MIMIC-III v1.4 database (2), and a lifelogging dataset named PMData (3).

Results

Three pivotal dimensions were assessed on the generated synthetic data: resemblance to the original data (1), utility (2), and privacy level (3). The optimal approach fluctuates based on the assessed dimension and metric.

Conclusion

The initial characteristics of the datasets to be synthesized play a crucial role in determining the best approach. Coupling synthetic metadata with real time series (A1), as well as jointly generating synthetic time series and metadata (A3), are both competitive methods, while separately generating time series and metadata (A2) appears to perform more poorly overall.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

MetadataPrivacyHumansTime FactorsDatabases, Factual

Stichwörter der Autor:innen

Health dataPrivacy-preserving data sharingSynthetic dataTime series

Autor:innen und Institutionen

  • Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA), Donostia-San Sebastián, Spain.
  • Computer Science and Artificial Intelligence Department, Computer Science Faculty, University of the Basque Country (UPV/EHU), Donostia - San Sebastian, Spain.
  • Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA), Donostia-San Sebastián, Spain. gepelde@vicomtech.org.
  • eHealth Group, Biogipuzkoa Health Research Institute, Donostia-San Sebastian, Spain. gepelde@vicomtech.org.
  • eHealth Group, Biogipuzkoa Health Research Institute, Donostia-San Sebastian, Spain.
  • Biomedical Engineering Department, Mondragon University, Arrasate-Mondragon, Spain.

Förderung

  • Horizon 2020 Framework Programme
  • Department of Education, Universities and Research of the Basque Country

Zitieren

APA

Isasa, I., Hernandez, M., Epelde, G., Londoño, F., Beristain, A., Larrea, X., Alberdi, A., Bamidis, P., & Konstantinidis, E. (2024). Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis. BMC Medical Informatics and Decision Making. https://doi.org/10.1186/s12911-024-02427-0

Vancouver

Isasa I, Hernandez M, Epelde G, Londoño F, Beristain A, Larrea X, et al. Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis. BMC Medical Informatics and Decision Making. 2024. doi:10.1186/s12911-024-02427-0

BibTeX
@article{Isasa2024,
  author = {Isasa, Imanol and Hernandez, Mikel and Epelde, Gorka and Londoño, Francisco and Beristain, Andoni and Larrea, Xabat and Alberdi, Ane and Bamidis, Panagiotis and Konstantinidis, Evdokimos},
  title = {Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis},
  journal = {BMC Medical Informatics and Decision Making},
  year = {2024},
  doi = {10.1186/s12911-024-02427-0}
}
Kostenlos starten →