Medical data wrangling with sequential variational autoencoders

Abstract

Medical data sets are usually corrupted by noise and missing data. These missing patterns are commonly assumed to be completely random, but in medical scenarios, the reality is that these patterns occur in bursts due to sensors that are off for some time or data collected in a misaligned uneven fashion, among other causes. This paper proposes to model medical data records with heterogeneous data types and bursty missing data using sequential variational autoencoders (VAEs). In particular, we propose a new methodology, the Shi-VAE, which extends the capabilities of VAEs to sequential streams of data with missing observations. We compare our model against state-of-the-art solutions in an intensive care unit database (ICU) and a dataset of passive human monitoring. Furthermore, we find that standard error metrics such as RMSE are not conclusive enough to assess temporal models and include in our analysis the cross-correlation between the ground truth and the imputed signal. We show that Shi-VAE achieves the best performance in terms of using both metrics, with lower computational complexity than the GP-VAE model, which is the state-of-the-art method for medical records.

arXiv commentAccepted for publication in IEEE Journal of Biomedical and Health Informatics (JBHI)

References in corpus (13)

• Scikit-learn: Machine Learning in Python
• Auto-Encoding Variational Bayes
• Categorical Reparameterization with Gumbel-Softmax
• A Recurrent Latent Variable Model for Sequential Data
• GAIN: Missing Data Imputation using Generative Adversarial Nets
• Learning to Diagnose with LSTM Recurrent Neural Networks
• Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoders
• BRITS: Bidirectional Recurrent Imputation for Time Series
• Modeling Missing Data in Clinical Time Series with RNNs
• MisGAN: Learning from Incomplete Data with Generative Adversarial Networks
• EDDI: Efficient Dynamic Discovery of High-Value Information with Partial VAE
• VAEM: a Deep Generative Model for Heterogeneous Mixed Type Data
• VAEs in the Presence of Missing Data