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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent work in machine learning for healthcare has raised concerns aboutpatient privacy and algorithmic fairness. For example, previous work has shownthat patient self-reported race can be predicted from medical data that doesnot explicitly contain racial information. However, the extent of dataidentification is unknown, and we lack ways to develop models whose outcomesare minimally affected by such information. Here we systematically investigatedthe ability of time-series electronic health record data to predict patientstatic information. We found that not only the raw time-series data, but alsolearned representations from machine learning models, can be trained to predicta variety of static information with area under the receiver operatingcharacteristic curve as high as 0.851 for biological sex, 0.869 for binarizedage and 0.810 for self-reported race. Such high predictive performance can beextended to a wide range of comorbidity factors and exists even when the modelwas trained for different tasks, using different cohorts, using different modelarchitectures and databases. Given the privacy and fairness concerns thesefindings pose, we develop a variational autoencoder-based approach that learnsa structured latent space to disentangle patient-sensitive attributes fromtime-series data. Our work thoroughly investigates the ability of machinelearning models to encode patient static information from time-serieselectronic health records and introduces a general approach to protectpatient-sensitive attribute information for downstream tasks.

Learning and DiSentangling Patient Static Information from Time-series Electronic HEalth Record (STEER)