Bayesian Double Feature Allocation for Phenotyping With Electronic Health Records

DOI10.1080/01621459.2019.1686985zbMATH Open1452.62835arXiv1809.08988OpenAlexW2982606431WikidataQ126856223 ScholiaQ126856223MaRDI QIDQ5146013FDOQ5146013

Authors: Yang Ni, P. Müller, Yuan Ji 

Publication date: 22 January 2021

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: We propose a categorical matrix factorization method to infer latent diseases from electronic health records (EHR) data in an unsupervised manner. A latent disease is defined as an unknown biological aberration that causes a set of common symptoms for a group of patients. The proposed approach is based on a novel double feature allocation model which simultaneously allocates features to the rows and the columns of a categorical matrix. Using a Bayesian approach, available prior information on known diseases greatly improves identifiability and interpretability of latent diseases. This includes known diagnoses for patients and known association of diseases with symptoms. We validate the proposed approach by simulation studies including mis-specified models and comparison with sparse latent factor models. In the application to Chinese EHR data, we find interesting results, some of which agree with related clinical and medical knowledge.

Full work available at URL: https://arxiv.org/abs/1809.08988

Recommendations

• A two‐phase Bayesian methodology for the analysis of binary phenotypes in genome‐wide association studies
• Bayesian latent multi‐state modeling for nonequidistant longitudinal electronic health records
• Bayesian analysis for imbalanced positive-unlabelled diagnosis codes in electronic health records
• Statistical inference for association studies using electronic health records: handling both selection bias and outcome misclassification
• Electronic health record analysis via deep Poisson factor models
• Bayesian and Frequentist Methods for Provider Profiling Using Risk-Adjusted Assessments of Medical Outcomes
• Bayesian Variable Selection in Multinomial Probit Models to Identify Molecular Signatures of Disease Stage
• Bayesian feature allocation models for tumor heterogeneity

zbMATH Keywords

matrix factorizationoverlapping clusteringIndian buffet processpatient-level inferencetripartite networks

Mathematics Subject Classification ID

Bayesian inference (62F15) Classification and discrimination; cluster analysis (statistical aspects) (62H30) Applications of statistics to biology and medical sciences; meta analysis (62P10)

Cites Work

• Modeling with normalized random measure mixture models
• Sparse Bayesian infinite factor models
• Title not available (Why is that?)
• Bayesian inference for general Gaussian graphical models with application to multivariate lattice data
• Title not available (Why is that?)
• The Indian buffet process: an introduction and review
• Bayesian Inference for Gene Expression and Proteomics
• Bayes and empirical-Bayes multiplicity adjustment in the variable-selection problem
• MCMC for normalized random measure mixture models
• Mixture models with a prior on the number of components
• Cluster and feature modeling from combinatorial stochastic processes
• Statistical analysis of \(Q\)-matrix based diagnostic classification models
• A Bayesian feature allocation model for tumor heterogeneity
• Nonparametric Bayesian bi-clustering for next generation sequencing count data
• Bayesian estimation of the DINA \(Q\) matrix
• Sampling decomposable graphs using a Markov chain on junction trees
• Exchangeable trait allocations

Cited In (10)

• A Bayesian approach to restricted latent class models for scientifically structured clustering of multivariate binary outcomes
• Hierarchical infinite factor models for improving the prediction of surgical complications for geriatric patients
• dfa
• Electronic health record analysis via deep Poisson factor models
• Comparison and Bayesian Estimation of Feature Allocations
• The attraction Indian buffet distribution
• A Joint MLE Approach to Large-Scale Structured Latent Attribute Analysis
• Biclustering via semiparametric Bayesian inference
• Automated feature selection of predictors in electronic medical records data
• Gaussian process regression and classification using international classification of disease codes as covariates

Uses Software

• R
• shiny
• BDgraph
• BNPdensity

This page was built for publication: Bayesian Double Feature Allocation for Phenotyping With Electronic Health Records

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5146013)