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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Building models for health prediction based on Electronic Health Records(EHR) has become an active research area. EHR patient journey data consists ofpatient time-ordered clinical events/visits from patients. Most existingstudies focus on modeling long-term dependencies between visits, withoutexplicitly taking short-term correlations between consecutive visits intoaccount, where irregular time intervals, incorporated as auxiliary information,are fed into health prediction models to capture latent progressive patterns ofpatient journeys. We present a novel deep neural network with four modules totake into account the contributions of various variables for health prediction:i) the Stacked Attention module strengthens the deep semantics in clinicalevents within each patient journey and generates visit embeddings, ii) theShort-Term Temporal Attention module models short-term correlations betweenconsecutive visit embeddings while capturing the impact of time intervalswithin those visit embeddings, iii) the Long-Term Temporal Attention modulemodels long-term dependencies between visit embeddings while capturing theimpact of time intervals within those visit embeddings, iv) and finally, theCoupled Attention module adaptively aggregates the outputs of Short-TermTemporal Attention and Long-Term Temporal Attention modules to make healthpredictions. Experimental results on MIMIC-III demonstrate superior predictiveaccuracy of our model compared to existing state-of-the-art methods, as well asthe interpretability and robustness of this approach. Furthermore, we foundthat modeling short-term correlations contributes to local priors generation,leading to improved predictive modeling of patient journeys.

Modeling Long-term Dependencies and Short-term Correlations in Patient Journey Data with Temporal Attention Networks for Health Prediction