Abstract:Abstract Keratitis is the main cause of corneal blindness worldwide. Most vision loss caused by keratitis can be avoidable via early detection and treatment. The diagnosis of keratitis often requires skilled ophthalmologists. However, the world is short of ophthalmologists, especially in resource-limited settings, making the early diagnosis of keratitis challenging. Here, we develop a deep learning system for the automated classification of keratitis, other cornea abnormalities, and normal cornea based on 6,567 slit-lamp images. Our system exhibits remarkable performance in cornea images captured by the different types of digital slit lamp cameras and a smartphone with the super macro mode (all AUCs>0.96). The comparable sensitivity and specificity in keratitis detection are observed between the system and experienced cornea specialists. Our system has the potential to be applied to both digital slit lamp cameras and smartphones to promote the early diagnosis and treatment of keratitis, preventing the corneal blindness caused by keratitis.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is corneal blindness caused by keratitis. Specifically, the goal of the paper is to develop a deep - learning - based system for automatically classifying keratitis, other corneal abnormalities, and normal corneas, thereby promoting the early diagnosis and treatment of keratitis and preventing blindness caused by keratitis. The following are the main contents and achievements of the paper: ### Research Background - **Keratitis** is the main cause of corneal blindness worldwide, especially in areas with limited resources. Due to the lack of ophthalmologists, early diagnosis and treatment become very difficult. - **Importance of Early Diagnosis**: Early detection and timely treatment can stop the progression of the disease, improve the prognosis, protect vision, and even maintain the integrity of the eyeball. If not treated early, keratitis may rapidly deteriorate, leading to permanent vision loss or even corneal perforation. ### Research Methods - **Data Set**: The research team used 6,567 slit - lamp images from different medical institutions, as well as images taken in the super - macro mode of smartphones. - **Deep - Learning Algorithms**: Three classic deep - learning algorithms, DenseNet121, Inception - v3, and ResNet50, were used in the study to train the model for classification. - **Performance Evaluation**: The performance of the system was evaluated through internal and external test sets (including three slit - lamp image data sets and one smartphone image data set). ### Main Results - **Internal Test Set**: - The DenseNet121 algorithm performed best in detecting keratitis, with an AUC value of 0.998, a sensitivity of 97.7%, and a specificity of 98.2%. - For the detection of other corneal abnormalities, the AUC value was 0.994, the sensitivity was 94.6%, and the specificity was 98.4%. - For the detection of normal corneas, the AUC value was 0.999, the sensitivity was 98.4%, and the specificity was 99.8%. - **External Test Set**: - In the external test set, the DenseNet121 algorithm also performed excellently, with AUC values ranging from 0.967 to 0.997. - The system also performed very well on smartphone images, with an AUC value of 0.967, a sensitivity of 91.9%, and a specificity of 96.9%. ### Discussion - **Advantages of the System**: - The system performs well on different types of slit - lamp cameras and smartphones and has wide applicability. - The performance of the system is comparable to that of experienced corneal experts and can be used as an early screening tool, especially in areas with limited resources. - By generating heat maps, the system can explain its decision - making process, enhancing its interpretability and credibility in practical applications. - **Limitations**: - The use of two - dimensional images may lead to some misclassifications, especially when the images lack a sense of three - dimensionality. - The system cannot make a specific diagnosis based solely on images and needs to be combined with other clinical information. - An image quality control system has not been developed to detect and filter low - quality images, which may affect the subsequent AI diagnosis system. ### Conclusion The deep - learning system developed in this study performs excellently in detecting keratitis and other corneal abnormalities and has broad application prospects. Especially in areas with limited resources, it can help in the early detection and treatment of keratitis and prevent corneal blindness.

Preventing corneal blindness caused by keratitis using artificial intelligence

Automatic Diagnosis of Keratitis Using Object Localization Combined with Cost-Sensitive Deep Attention Convolutional Neural Network

The application and value evaluation of assisted diagnosis system for five fundus lesion based on artificial intelligence combined with optical coherence tomography

Deep learning for multi-type infectious keratitis diagnosis: A nationwide, cross-sectional, multicenter study

Enabling Collaborative Clinical Diagnosis of Infectious Keratitis by Integrating Expert Knowledge and Interpretable Data-driven Intelligence

Feasibility assessment of infectious keratitis depicted on slit-lamp and smartphone photographs using deep learning

Automatic Diagnosis of Infectious Keratitis Based on Slit Lamp Images Analysis

Establishment of an Automatic Diagnosis System for Corneal Endothelium Diseases Using Artificial Intelligence

Promoting smartphone-based keratitis screening using meta-learning: A multicenter study

Deep learning-based classification of infectious keratitis on slit-lamp images

Deep learning model for extensive smartphone-based diagnosis and triage of cataracts and multiple corneal diseases

DeepMonitoring: a deep learning-based monitoring system for assessing the quality of cornea images captured by smartphones

Determination of probability of causative pathogen in infectious keratitis using deep learning algorithm of slit-lamp images

Intelligent Diagnosis and Classification of Keratitis

Class-Aware Attention Network for infectious keratitis diagnosis using corneal photographs

Deep learning for identifying corneal diseases from ocular surface slit-lamp photographs

Artificial intelligence in the anterior segment of eye diseases

Development and validation of a deep learning system to screen vision-threatening conditions in high myopia using optical coherence tomography images

Diagnostic performance of deep learning for infectious keratitis: a systematic review and meta-analysis

[Deep learning based lesion detection from anterior segment optical coherence tomography images and its application in the diagnosis of keratoconus].

Artificial intelligence in corneal diseases: A narrative review