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

In this paper, we consider the problem of disease diagnosis. Unlike theconventional learning paradigm that treats labels independently, we propose aknowledge-enhanced framework, that enables training visual representation withthe guidance of medical domain knowledge. In particular, we make the followingcontributions: First, to explicitly incorporate experts' knowledge, we proposeto learn a neural representation for the medical knowledge graph viacontrastive learning, implicitly establishing relations between differentmedical concepts. Second, while training the visual encoder, we keep theparameters of the knowledge encoder frozen and propose to learn a set of promptvectors for efficient adaptation. Third, we adopt a Transformer-baseddisease-query module for cross-model fusion, which naturally enablesexplainable diagnosis results via cross attention. To validate theeffectiveness of our proposed framework, we conduct thorough experiments onthree x-ray imaging datasets across different anatomy structures, showing ourmodel is able to exploit the implicit relations between diseases/findings, thusis beneficial to the commonly encountered problem in the medical domain,namely, long-tailed and zero-shot recognition, which conventional methodseither struggle or completely fail to realize.

K-Diag: Knowledge-enhanced Disease Diagnosis in Radiographic Imaging