Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images

Summary: The performance of deep learning in disease detection from high-quality clinical images is identical to and even greater than that of human doctors. However, in low-quality images, deep learning performs poorly. Whether human doctors also have poor performance in low-quality images is unkno...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Zhongwen Li, Jiewei Jiang, Wei Qiang, Liufei Guo, Xiaotian Liu, Hongfei Weng, Shanjun Wu, Qinxiang Zheng, Wei Chen
Formato:	article
Lenguaje:	EN
Publicado:	Elsevier 2021
Materias:	Ocular surface Ophthalmology Artificial intelligence Science Q
Acceso en línea:	https://doaj.org/article/9b93422a7e1e42fcba8ab69db7af6fd6
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:9b93422a7e1e42fcba8ab69db7af6fd6
record_format	dspace
spelling	oai:doaj.org-article:9b93422a7e1e42fcba8ab69db7af6fd62021-11-20T05:09:54ZComparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images2589-004210.1016/j.isci.2021.103317https://doaj.org/article/9b93422a7e1e42fcba8ab69db7af6fd62021-11-01T00:00:00Zhttp://www.sciencedirect.com/science/article/pii/S2589004221012864https://doaj.org/toc/2589-0042Summary: The performance of deep learning in disease detection from high-quality clinical images is identical to and even greater than that of human doctors. However, in low-quality images, deep learning performs poorly. Whether human doctors also have poor performance in low-quality images is unknown. Here, we compared the performance of deep learning systems with that of cornea specialists in detecting corneal diseases from low-quality slit lamp images. The results showed that the cornea specialists performed better than our previously established deep learning system (PEDLS) trained on only high-quality images. The performance of the system trained on both high- and low-quality images was superior to that of the PEDLS while inferior to that of a senior corneal specialist. This study highlights that cornea specialists perform better in low-quality images than the system trained on high-quality images. Adding low-quality images with sufficient diagnostic certainty to the training set can reduce this performance gap.Zhongwen LiJiewei JiangWei QiangLiufei GuoXiaotian LiuHongfei WengShanjun WuQinxiang ZhengWei ChenElsevierarticleOcular surfaceOphthalmologyArtificial intelligenceScienceQENiScience, Vol 24, Iss 11, Pp 103317- (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Ocular surface Ophthalmology Artificial intelligence Science Q
spellingShingle	Ocular surface Ophthalmology Artificial intelligence Science Q Zhongwen Li Jiewei Jiang Wei Qiang Liufei Guo Xiaotian Liu Hongfei Weng Shanjun Wu Qinxiang Zheng Wei Chen Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
description	Summary: The performance of deep learning in disease detection from high-quality clinical images is identical to and even greater than that of human doctors. However, in low-quality images, deep learning performs poorly. Whether human doctors also have poor performance in low-quality images is unknown. Here, we compared the performance of deep learning systems with that of cornea specialists in detecting corneal diseases from low-quality slit lamp images. The results showed that the cornea specialists performed better than our previously established deep learning system (PEDLS) trained on only high-quality images. The performance of the system trained on both high- and low-quality images was superior to that of the PEDLS while inferior to that of a senior corneal specialist. This study highlights that cornea specialists perform better in low-quality images than the system trained on high-quality images. Adding low-quality images with sufficient diagnostic certainty to the training set can reduce this performance gap.
format	article
author	Zhongwen Li Jiewei Jiang Wei Qiang Liufei Guo Xiaotian Liu Hongfei Weng Shanjun Wu Qinxiang Zheng Wei Chen
author_facet	Zhongwen Li Jiewei Jiang Wei Qiang Liufei Guo Xiaotian Liu Hongfei Weng Shanjun Wu Qinxiang Zheng Wei Chen
author_sort	Zhongwen Li
title	Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
title_short	Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
title_full	Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
title_fullStr	Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
title_full_unstemmed	Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
title_sort	comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
publisher	Elsevier
publishDate	2021
url	https://doaj.org/article/9b93422a7e1e42fcba8ab69db7af6fd6
work_keys_str_mv	AT zhongwenli comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT jieweijiang comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT weiqiang comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT liufeiguo comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT xiaotianliu comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT hongfeiweng comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT shanjunwu comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT qinxiangzheng comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages AT weichen comparisonofdeeplearningsystemsandcorneaspecialistsindetectingcornealdiseasesfromlowqualityimages
_version_	1718419557268324352

Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images

Ejemplares similares