A method for generating synthetic longitudinal health data

Mosquera, Lucy; El Emam, Khaled; Ding, Lei; Sharma, Vishal; Zhang, Xue Hua; Kababji, Samer El; Carvalho, Chris; Hamilton, Brian; Palfrey, Dan; Kong, Linglong; Jiang, Bei; Eurich, Dean T.

doi:10.1186/s12874-023-01869-w

BMC Medical Research Methodology

Table 1 Literature review of key characteristics of previous works for generating longitudinal synthetic health data

From: A method for generating synthetic longitudinal health data

Title	Data Structure			Variable Types					Model Types
Title	Cross-sectional R.1a)	Longitudinal (R.1b)	Variable length sequences (R.2)	Categorical (R.3a)	Continuous (R.3b)	Categories with high cardinality (R.3c)	Outliers removed (R.4)	Missing values present in data (R.5)	Consider all the previous information (R.6)	Model informed by clinicians (R.7)
Variational Autoencoder Modular Bayesian Networks (VAMBN) for Simulation of Heterogeneous Clinical Study Data [61]	No	Yes	Fixed	Yes	Yes	Yes	N/D	Yes	Yes	No
Machine learning for comprehensive forecasting of Alzheimer’s Disease progression [62]	No	Yes	Varied	Yes	Yes	No	N/D	Yes	No	No
Design and Validation of a Data Simulation Model for Longitudinal Healthcare Data [63]	No	Yes	Varied	Yes	No	Yes	N/D	No	Yes	No
Privacy-Preserving Generative Deep Neural Networks Support Clinical Data Sharing [64]	No	Yes	Fixed	No	Yes	No	Yes	No	Yes	No
Analyzing Medical Research Results Based on Synthetic Data and Their Relation to Real Data Results: Systematic Comparison From Five Observational Studies [65]	Yes	No	N/A	Yes	Yes	No	N/D	Yes	N/A	No
Synthetic Event Time Series Health Data Generation [66]	Yes	Yes	Fixed	Yes	Yes	No	Yes	No	Yes	No
Data-driven approach for creating synthetic electronic medical records [67]	No	Yes	Varied	Yes	Yes	Yes	N/D	N/D	Yes	No
Synthea: An approach, method, and software mechanism for generating synthetic patients and the synthetic electronic health care record [68]	Yes	Yes	Varied	Yes	Yes	Yes	N/D	No	Yes	Yes
Real-valued (medical) time series generation with recurrent conditional GANS [69]	No	Yes	Fixed	No	Yes	N/A	Yes	No	Yes	No
Generating Multi-label Discrete Patient Records using Generative Adversarial Networks [70]	Yes	No	N/A	Yes	No	No	N/D	Yes	No	No
Data Synthesis based on Generative Adversarial Networks [35]	Yes	Yes	Fixed	Yes	Yes	Yes	N/D	N/D	Yes	No
Generation and Evaluation of Privacy Preserving Synthetic Health Data [71]	Yes	No	N/A	Yes	Yes	Yes	No	No	No	No
Generation of Heterogeneous Synthetic Electronic Health Records using GANs [72]	Yes	No	N/A	Yes	Yes	Yes	Yes	N/D	No	No
Generating Electronic Health Records with Multiple Data Types and Constraints [73]	Yes	No	N/A	Yes	Yes	Yes	Yes	N/D	No	No
Ensuring electronic medical record simulation through better training, modeling, and evaluation [74]	Yes	No	N/A	Yes	No	Yes	Yes	N/D	No	No
Generative Adversarial Networks for Electronic Health Records: A Framework for Exploring and Evaluating Methods for Predicting Drug-Induced Laboratory Test Trajectories [75]	No	Yes	Fixed	No	Yes	N/A	Yes	No	Yes	No
Synthesizing electronic health records using improved generative adversarial networks [76]	Yes	No	N/A	Yes	No	No	Yes	N/D	Yes	No
Generating Fake Data Using GANs for Anonymizing Healthcare Data [77]	Yes	Yes	Fixed	Yes	Yes	No	Yes	N/D	No	No
CorGAN: Correlation-Capturing Convolutional Generative Adversarial Networks for Generating Synthetic Healthcare Records [78]	Yes	No	N/A	Yes	Yes	No	N/D	N/D	N/A	No
Generation and evaluation of synthetic patient data [79]	Yes	No	N/A	Yes	Yes	No	No	N/D	N/A	No
Generating and Evaluating Synthetic UK Primary Care Data: Preserving Data Utility & Patient Privacy [80]	Yes	No	N/A	Yes	Yes	No	No	N/D	N/A	No
SMOOTH-GAN: Towards Sharp and Smooth Synthetic EHR Data Generation [81]	Yes	No	N/A	Yes	Yes	No	Yes	N/D	N/A	No
Continuous Patient-Centric Sequence Generation via Sequentially Coupled Adversarial Learning [82]	No	Yes	Varied	No	Yes	N/A	Yes	No	Yes	No
Medical Time-Series Data Generation using Generative Adversarial Networks [83]	No	Yes	Varied	Yes	Yes	No	N/D	N/D	No	No

N/A refers to not applicable while N/D refers to not described

Back to article page

ISSN: 1471-2288

Contact us

General enquiries: journalsubmissions@springernature.com