Comparison of deep learning systems and cornea specialists in detecting corneal diseases from low-quality images
Zhongwen Li,
Jiewei Jiang,
Wei Qiang,
Liufei Guo,
Xiaotian Liu,
Hongfei Weng,
Shanjun Wu,
Qinxiang Zheng,
Wei Chen
Affiliations
Zhongwen Li
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China; School of Ophthalmology and Optometry and Eye Hospital, Wenzhou Medical University, Wenzhou 325027, China
Jiewei Jiang
School of Electronic Engineering, Xi'an University of Posts and Telecommunications, Xi'an 710121, China
Wei Qiang
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China
Liufei Guo
School of Electronic Engineering, Xi'an University of Posts and Telecommunications, Xi'an 710121, China
Xiaotian Liu
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China
Hongfei Weng
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China
Shanjun Wu
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China
Qinxiang Zheng
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China; School of Ophthalmology and Optometry and Eye Hospital, Wenzhou Medical University, Wenzhou 325027, China; Corresponding author
Wei Chen
Ningbo Eye Hospital, Wenzhou Medical University, Ningbo 315000, China; School of Ophthalmology and Optometry and Eye Hospital, Wenzhou Medical University, Wenzhou 325027, China; Corresponding author
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.
