Log in to save to my catalogue

MDKLoss: Medicine domain knowledge loss for skin lesion recognition

MDKLoss: Medicine domain knowledge loss for skin lesion recognition

https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_a7d9cbe4a3b94613b3d782d3139e1ff2

MDKLoss: Medicine domain knowledge loss for skin lesion recognition

About this item

Full title

MDKLoss: Medicine domain knowledge loss for skin lesion recognition

Publisher

United States: AIMS Press

Journal title

Mathematical biosciences and engineering : MBE, 2024, Vol.21 (2), p.2671-2690

Language

English

Formats

Publication information

Publisher

United States: AIMS Press

More information

Scope and Contents

Contents

Methods based on deep learning have shown good advantages in skin lesion recognition. However, the diversity of lesion shapes and the influence of noise disturbances such as hair, bubbles, and markers leads to large intra-class differences and small inter-class similarities, which existing methods have not yet effectively resolved. In addition, mos...

Alternative Titles

Full title

MDKLoss: Medicine domain knowledge loss for skin lesion recognition

Authors, Artists and Contributors

Identifiers

Primary Identifiers

Record Identifier

TN_cdi_doaj_primary_oai_doaj_org_article_a7d9cbe4a3b94613b3d782d3139e1ff2

Permalink

https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_a7d9cbe4a3b94613b3d782d3139e1ff2

Other Identifiers

ISSN

1551-0018

E-ISSN

1551-0018

DOI

10.3934/mbe.2024118

How to access this item