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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Beyond generating long and topic-coherent paragraphs in traditionalcaptioning tasks, the medical image report composition task poses moretask-oriented challenges by requiring both the highly-accurate medical termdiagnosis and multiple heterogeneous forms of information including impressionand findings. Current methods often generate the most common sentences due todataset bias for individual case, regardless of whether the sentences properlycapture key entities and relationships. Such limitations severely hinder theirapplicability and generalization capability in medical report composition wherethe most critical sentences lie in the descriptions of abnormal diseases thatare relatively rare. Moreover, some medical terms appearing in one report areoften entangled with each other and co-occurred, e.g. symptoms associated witha specific disease. To enforce the semantic consistency of medical terms to beincorporated into the final reports and encourage the sentence generation forrare abnormal descriptions, we propose a novel framework that unifies templateretrieval and sentence generation to handle both common and rare abnormalitywhile ensuring the semantic-coherency among the detected medical terms.Specifically, our approach exploits hybrid-knowledge co-reasoning: i) explicitrelationships among all abnormal medical terms to induce the visual attentionlearning and topic representation encoding for better topic-oriented symptomsdescriptions; ii) adaptive generation mode that changes between the templateretrieval and sentence generation according to a contextual topic encoder.Experimental results on two medical report benchmarks demonstrate thesuperiority of the proposed framework in terms of both human and metricsevaluation.

Unifying Relational Sentence Generation and Retrieval for Medical Image Report Composition