Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis
Abstract 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...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2024-01-01
|
| Series: | BMC Medical Informatics and Decision Making |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12911-024-02427-0 |
| _version_ | 1797274416076292096 |
|---|---|
| author | Imanol Isasa Mikel Hernandez Gorka Epelde Francisco Londoño Andoni Beristain Xabat Larrea Ane Alberdi Panagiotis Bamidis Evdokimos Konstantinidis |
| author_facet | Imanol Isasa Mikel Hernandez Gorka Epelde Francisco Londoño Andoni Beristain Xabat Larrea Ane Alberdi Panagiotis Bamidis Evdokimos Konstantinidis |
| author_sort | Imanol Isasa |
| collection | DOAJ |
| description | Abstract 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. |
| first_indexed | 2024-03-07T14:58:00Z |
| format | Article |
| id | doaj.art-fa415233cda846b5bf8b77eeeaf389b0 |
| institution | Directory Open Access Journal |
| issn | 1472-6947 |
| language | English |
| last_indexed | 2024-03-07T14:58:00Z |
| publishDate | 2024-01-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Informatics and Decision Making |
| spelling | doaj.art-fa415233cda846b5bf8b77eeeaf389b02024-03-05T19:20:01ZengBMCBMC Medical Informatics and Decision Making1472-69472024-01-0124111410.1186/s12911-024-02427-0Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesisImanol Isasa0Mikel Hernandez1Gorka Epelde2Francisco Londoño3Andoni Beristain4Xabat Larrea5Ane Alberdi6Panagiotis Bamidis7Evdokimos Konstantinidis8Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA)Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA)Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA)Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA)Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA)Digital Health and Biomedical Technologies, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA)Biomedical Engineering Department, Mondragon UniversityLaboratory of Medical Physics and Digital Innovation, School of Medicine, Aristotle University of ThessalonikiLaboratory of Medical Physics and Digital Innovation, School of Medicine, Aristotle University of ThessalonikiAbstract 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.https://doi.org/10.1186/s12911-024-02427-0Time seriesSynthetic dataPrivacy-preserving data sharingHealth data |
| spellingShingle | Imanol Isasa Mikel Hernandez Gorka Epelde Francisco Londoño Andoni Beristain Xabat Larrea Ane Alberdi Panagiotis Bamidis Evdokimos Konstantinidis Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis BMC Medical Informatics and Decision Making Time series Synthetic data Privacy-preserving data sharing Health data |
| title | Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis |
| title_full | Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis |
| title_fullStr | Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis |
| title_full_unstemmed | Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis |
| title_short | Comparative assessment of synthetic time series generation approaches in healthcare: leveraging patient metadata for accurate data synthesis |
| title_sort | comparative assessment of synthetic time series generation approaches in healthcare leveraging patient metadata for accurate data synthesis |
| topic | Time series Synthetic data Privacy-preserving data sharing Health data |
| url | https://doi.org/10.1186/s12911-024-02427-0 |
| work_keys_str_mv | AT imanolisasa comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT mikelhernandez comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT gorkaepelde comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT franciscolondono comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT andoniberistain comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT xabatlarrea comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT anealberdi comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT panagiotisbamidis comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis AT evdokimoskonstantinidis comparativeassessmentofsynthetictimeseriesgenerationapproachesinhealthcareleveragingpatientmetadataforaccuratedatasynthesis |