Retracted: Deep Learning-Based Glaucoma Detection Using CNN and Digital Fundus Images: A Promising Approach for...

Area of Science:

• Ophthalmology
• Artificial Intelligence
• Medical Imaging

Background:

• Glaucoma is a leading cause of irreversible blindness globally.
• Early symptoms are often undetected, leading to significant vision loss.
• Current diagnostic methods have limitations, including late detection and reliance on subjective feedback.

Purpose of the Study:

• To develop an AI-based method for glaucoma detection.
• To reduce glaucoma-related blindness through precise diagnosis.
• To overcome limitations of existing glaucoma screening techniques.

Main Methods:

• Utilized Heidelberg Retinal Tomography (HRT), Optical Coherence Tomography (OCT), and Fundus Photography.
• Employed Support Vector Machines (SVMs) and Convolutional Neural Networks (CNNs) for analysis.
• Analyzed 20 fundus images from the RIM-ONE-r3 dataset (Healthy, Glaucoma, Suspects).

Main Results:

• The AI model demonstrated high diagnostic accuracy in glaucoma detection.
• Fundus image recognition showed promising outcomes on the RIM-ONE-r3 dataset.
• Consistently high accuracy rates were achieved across image categories.

Conclusions:

• Augmenting datasets with more labeled images can enhance AI model accuracy.
• Integration of computer-aided systems requires careful consideration of application parameters.
• The study highlights the potential of AI in improving glaucoma diagnosis and prevention.