Generating a Glaucoma Diagnosis Report Using Deep Learning and Humphrey Visual Field Data

Glaucoma remains a leading cause of irreversible blindness worldwide, emphasizing the need for early and accurate diagnosis.This study presents an automated system for evaluating visual field data using the Humphrey Field Analyzer.By integrating deep learning with Optical Character Recognition (OCR), the proposed model extracts critical clinical parameters from visual field reports, processes them through a trained neural network, and generates structured diagnostic reports.The system was trained on a dataset of Humphrey Visual Field (HVF) images, where key features such as Age, Central 5 threshold values, Mean Deviation (MD), and Pattern Standard Deviation (PSD) were used for classification.Experimental results demonstrated that the proposed model achieved an accuracy of 97.8%, surpassing both traditional manual interpretation (85%) and convolutional neural network-based image classification (93.7%).The system enhances diagnostic consistency and reduces interobserver variability, making it a reliable alternative to conventional methods.However, its performance is influenced by OCR accuracy and variations in test conditions, which may introduce errors in data extraction.The findings highlight the potential of AI-driven systems in clinical ophthalmology, offering a scalable and efficient approach for automated glaucoma assessment and personalized treatment planning.