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

The computer-aided detection (CADe) systems are developed to assistpathologists in slide assessment, increasing diagnosis efficiency and reducingmissing inspections. Many studies have shown such a CADe system with deeplearning approaches outperforms the one using conventional methods that rely onhand-crafted features based on field-knowledge. However, most developers whoadopted deep learning models directly focused on the efficacy of outcomes,without providing comprehensive explanations on why their proposed frameworkscan work effectively. In this study, we designed four experiments to verify theconsecutive concepts, showing that the deep features learned from pathologicalpatches are interpretable by domain knowledge of pathology and enlightening forclinical diagnosis in the task of lesion detection. The experimental resultsshow the activation features work as morphological descriptors for specificcells or tissues, which agree with the clinical rules in classification. Thatis, the deep learning framework not only detects the distribution of tumorcells but also recognizes lymphocytes, collagen fibers, and some other non-cellstructural tissues. Most of the characteristics learned by the deep learningmodels have summarized the detection rules that can be recognized by theexperienced pathologists, whereas there are still some features may not beintuitive to domain experts but discriminative in classification for machines.Those features are worthy to be further studied in order to find out thereasonable correlations to pathological knowledge, from which pathologicalexperts may draw inspirations for exploring new characteristics in diagnosis.

Understanding the Mechanism of Deep Learning Framework for Lesion Detection in Pathological Images with Breast Cancer