Deep-learning-based, computer-aided classifier developed with a small dataset of clinical images surpasses board-certified dermatologists in skin tumour diagnosis.

Journal: The British journal of dermatology

Published Date: Feb 1, 2019

Abstract

BACKGROUND: Application of deep-learning technology to skin cancer classification can potentially improve the sensitivity and specificity of skin cancer screening, but the number of training images required for such a system is thought to be extremely large.

Authors

Y Fujisawa

Dermatology Division, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, Japan, 305-8577.
Y Otomo

Kyocera Communications System Co., Ltd, Kyoto, Japan.
Y Ogata

KCCS Mobile Engineering Co., Ltd, Tokyo, Japan.
Y Nakamura

Dermatology Division, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, Japan, 305-8577.
R Fujita

Kyocera Communications System Co., Ltd, Kyoto, Japan.
Y Ishitsuka

Dermatology Division, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, Japan, 305-8577.
R Watanabe

Dermatology Division, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, Japan, 305-8577.
N Okiyama

Dermatology Division, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, Japan, 305-8577.
K Ohara

Dermatology, Akasaka Toranomon Clinic, Tokyo, Japan.
M Fujimoto

Dermatology Division, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki, Japan, 305-8577.

Keywords

Datasets as Topic Deep Learning Dermatologists Dermoscopy Humans Image Interpretation, Computer-Assisted Mobile Applications Sensitivity and Specificity Skin Skin Neoplasms Smartphone

External Resources

View on PubMed Access via DOI PubMed (29953582)

Deep-learning-based, computer-aided classifier developed with a small dataset of clinical images surpasses board-certified dermatologists in skin tumour diagnosis.

Abstract

Authors

Keywords

External Resources

Popular Topics

Recent Journals