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/

Zendy Open

Presents the pdf analysis as premium feature

PDF Analysis

Insights

Presents the summarisation as premium feature

Summarisation

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Sepsis is the leading cause of in-hospital mortality in the USA. Early sepsisonset prediction and diagnosis could significantly improve the survival ofsepsis patients. Existing predictive models are usually trained on high-qualitydata with few missing information, while missing values widely exist inreal-world clinical scenarios (especially in the first hours of admissions tothe hospital), which causes a significant decrease in accuracy and an increasein uncertainty for the predictive models. The common method to handle missingvalues is imputation, which replaces the unavailable variables with estimatesfrom the observed data. The uncertainty of imputation results can be propagatedto the sepsis prediction outputs, which have not been studied in existing workson either sepsis prediction or uncertainty quantification. In this study, wefirst define such propagated uncertainty as the variance of prediction outputand then introduce uncertainty propagation methods to quantify the propagateduncertainty. Moreover, for the potential high-risk patients with low confidencedue to limited observations, we propose a robust active sensing algorithm toincrease confidence by actively recommending clinicians to observe the mostinformative variables. We validate the proposed models in both publiclyavailable data (i.e., MIMIC-III and AmsterdamUMCdb) and proprietary data in TheOhio State University Wexner Medical Center (OSUWMC). The experimental resultsshow that the propagated uncertainty is dominant at the beginning of admissionsto hospitals and the proposed algorithm outperforms state-of-the-art activesensing methods. Finally, we implement a SepsisLab system for early sepsisprediction and active sensing based on our pre-trained models. Clinicians andpotential sepsis patients can benefit from the system in early prediction anddiagnosis of sepsis.

SepsisLab: Early Sepsis Prediction with Uncertainty Quantification and Active Sensing