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

Decision support systems based on clinical notes have the potential toimprove patient care by pointing doctors towards overseen risks. Predicting apatient's outcome is an essential part of such systems, for which the use ofdeep neural networks has shown promising results. However, the patterns learnedby these networks are mostly opaque and previous work revealed flaws regardingthe reproduction of unintended biases. We thus introduce an extendable testingframework that evaluates the behavior of clinical outcome models regardingchanges of the input. The framework helps to understand learned patterns andtheir influence on model decisions. In this work, we apply it to analyse thechange in behavior with regard to the patient characteristics gender, age andethnicity. Our evaluation of three current clinical NLP models demonstrates theconcrete effects of these characteristics on the models' decisions. They showthat model behavior varies drastically even when fine-tuned on the same dataand that allegedly best-performing models have not always learned the mostmedically plausible patterns.

What Do You See in this Patient? Behavioral Testing of Clinical NLP Models