Variational Autoencoder Modular Bayesian Networks (VAMBN) for Simulation of Heterogeneous Clinical Study Data

In the area of Big Data one of the major obstacles for the progress of biomedical research is the existence of data “silos”, because legal and ethical constraints often do not allow for sharing sensitive patient data from clinical studies across institutions. While federated machine learning now allows for building models from scattered data, there is still the need to investigate, mine and understand clinical data that cannot be accessed directly. Simulation of sufficiently realistic virtual patients could be a way to fill this gap. In this work we propose a new machine learning approach (VAMBN) to learn a generative model of longitudinal clinical study data. VAMBN considers typical key aspects of such data, namely limited sample size coupled with comparable many variables of different numerical scales and statistical properties, and many missing values. We show that with VAMBN we can simulate virtual patients in a sufficiently realistic manner while making theoretical guarantees on data privacy. In addition, VAMBN allows for simulating counterfactual scenarios. Hence, VAMBN could facilitate data sharing as well as design of clinical trials. * VAMBN : Variational Autoencoder Modular Bayesian Network BN : Bayesian Network MBN : Modular Bayesian Network VAE : Variational Autoencoder HI-VAE : Heterogenous and Incomplete Data Variational Autoencoder DAG : directed acyclic graph MCAR : missing completely at random MAR : missing at random MNAR : missing not at random BIC : Bayesian Information Score PPMI : Parkinson’s Progression Marker Initiative PD : Parkinson’s Disease UPDRS : Unified Parkinson’s Disease Rating Scales ESS : Epworth Sleepiness Scale RBD : REM sleep behavior disorder CSF : cerebrospinal fluid