English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Machine learning has become ubiquitous and a key technology on miningelectronic health records (EHRs) for facilitating clinical research andpractice. Unsupervised machine learning, as opposed to supervised learning, hasshown promise in identifying novel patterns and relations from EHRs withoutusing human created labels. In this paper, we investigate the application ofunsupervised machine learning models in discovering latent disease clusters andpatient subgroups based on EHRs. We utilized Latent Dirichlet Allocation (LDA),a generative probabilistic model, and proposed a novel model named PoissonDirichlet Model (PDM), which extends the LDA approach using a Poissondistribution to model patients' disease diagnoses and to alleviate age and sexfactors by considering both observed and expected observations. In theempirical experiments, we evaluated LDA and PDM on three patient cohorts withEHR data retrieved from the Rochester Epidemiology Project (REP), for thediscovery of latent disease clusters and patient subgroups. We compared theeffectiveness of LDA and PDM in identifying latent disease clusters through thevisualization of disease representations learned by two approaches. We alsotested the performance of LDA and PDM in differentiating patient subgroupsthrough survival analysis, as well as statistical analysis. The experimentalresults show that the proposed PDM could effectively identify distinguisheddisease clusters by alleviating the impact of age and sex, and that LDA couldstratify patients into more differentiable subgroups than PDM in terms ofp-values. However, the subgroups discovered by PDM might imply the underlyingpatterns of diseases of greater interest in epidemiology research due to thealleviation of age and sex. Both unsupervised machine learning approaches couldbe leveraged to discover patient subgroups using EHRs but with different foci.

Unsupervised Machine Learning for the Discovery of Latent Disease Clusters and Patient Subgroups Using Electronic Health Records