Two-stage deep neural network for diagnosing fungal keratitis via in vivo confocal microscopy images

Li, Chun-Peng, Dai, Weiwei, Xiao, Yun-Peng, Qi, Mengying, Zhang, Ling-Xiao, Gao, Lin, Zhang, Fang-Lue, Lai, YuKun ORCID: https://orcid.org/0000-0002-2094-5680, Liu, Chang, Lu, Jing, Chen, Fen, Chen, Dan, Shi, Shuai, Li, Shaowei, Zeng, Qingyan and Chen, Yiqiang 2024. Two-stage deep neural network for diagnosing fungal keratitis via in vivo confocal microscopy images. Scientific Reports 14 (1) , 18432. 10.1038/s41598-024-68768-y

PDF - Published Version Download (2MB)

Abstract

Timely and effective diagnosis of fungal keratitis (FK) is necessary for suitable treatment and avoiding irreversible vision loss for patients. In vivo confocal microscopy (IVCM) has been widely adopted to guide the FK diagnosis. We present a deep learning framework for diagnosing fungal keratitis using IVCM images to assist ophthalmologists. Inspired by the real diagnostic process, our method employs a two-stage deep architecture for diagnostic predictions based on both image-level and sequence-level information. To the best of our knowledge, we collected the largest dataset with 96,632 IVCM images in total with expert labeling to train and evaluate our method. The specificity and sensitivity of our method in diagnosing FK on the unseen test set achieved 96.65% and 97.57%, comparable or better than experienced ophthalmologists. The network can provide image-level, sequence-level and patient-level diagnostic suggestions to physicians. The results show great promise for assisting ophthalmologists in FK diagnosis.

Item Type:	Article
Date Type:	Published Online
Status:	Published
Schools:	Computer Science & Informatics
Additional Information:	License information from Publisher: LICENSE 1: URL: http://creativecommons.org/licenses/by-nc-nd/4.0/, Type: open-access
Publisher:	Nature Research
Date of First Compliant Deposit:	12 August 2024
Date of Acceptance:	29 July 2024
Last Modified:	12 Aug 2024 10:16
URI:	https://orca.cardiff.ac.uk/id/eprint/171312

Actions (repository staff only)

Edit Item