AN ALGORITHM FOR IDENTIFYING SIGNS OF EYE DISEASES USING ARTIFICIAL INTELLIGENCE METHODS

Аннотация

The early detection of eye diseases plays a crucial role in preventing vision loss and improving treatment outcomes. With advancements in artificial intelligence (AI), particularly in machine learning and deep learning, it has become feasible to develop algorithms capable of identifying early signs of various eye conditions. This study presents an AI-based algorithm for detecting common eye diseases, such as diabetic retinopathy, glaucoma, and age-related macular degeneration (AMD). The proposed algorithm leverages convolutional neural networks (CNNs) to analyze medical imaging data, such as fundus photographs, optical coherence tomography (OCT), and visual field tests. Keywords. Artificial Intelligence, Eye Diseases, Diabetic Retinopathy, Glaucoma, Age-Related Macular Degeneration, Convolutional Neural Networks, Medical Imaging, Early Detection, Computer-Aided Diagnosis. Аннотация. Раннее выявление заболеваний глаз играет решающую роль в предотвращении потери зрения и улучшении результатов лечения. Благодаря достижениям в области искусственного интеллекта (ИИ), особенно в машинном обучении и глубоком обучении, стало возможным разрабатывать алгоритмы, способные определять ранние признаки различных заболеваний глаз. В этом исследовании представлен алгоритм на основе ИИ для выявления распространенных заболеваний глаз, таких как диабетическая ретинопатия, глаукома и возрастная макулярная дегенерация (ВМД). Предлагаемый алгоритм использует сверточные нейронные сети (CNN) для анализа данных медицинской визуализации, таких как фотографии глазного дна, оптическая когерентная томография (ОКТ) и тесты поля зрения. Выпуск журнала №-26 Часть–6_ Май –2025 92 MODERN EDUCATION AND DEVELOPMENT Ключевые слова. искусственный интеллект, заболевания глаз, диабетическая ретинопатия, глаукома, возрастная макулярная дегенерация, сверточные нейронные сети, медицинская визуализация, раннее выявление, компьютерная диагностика. Eye diseases are among the leading causes of vision impairment worldwide. The ability to detect these conditions at an early stage significantly enhances the effectiveness of treatments and helps prevent irreversible damage. Traditional diagnostic methods for eye diseases often require expert evaluation, which can be time- consuming and prone to human error. Recent advancements in artificial intelligence (AI) have opened new possibilities for automated diagnosis, offering faster, more accurate detection methods. Among these, machine learning and deep learning, particularly convolutional neural networks (CNNs), have shown promise in analyzing medical images and identifying disease markers. Numerous studies have explored the application of AI in the detection of eye diseases. Convolutional neural networks (CNNs) have been successfully used to classify retinal images for diabetic retinopathy detection. Similarly, other machine learning techniques have been employed to identify optic nerve abnormalities for glaucoma detection. Recent approaches have also focused on using OCT scans to detect macular degeneration, achieving high levels of accuracy in both classification and segmentation tasks. However, most of these methods require high-quality image datasets and substantial computational resources. Moreover, integrating different types of medical imaging (such as fundus photographs, OCT scans, and visual field tests) into a unified diagnostic algorithm remains a challenge. This paper aims to address these gaps by proposing a versatile AI-based algorithm that can handle various imaging modalities to detect multiple eye diseases. The proposed algorithm utilizes a convolutional neural network (CNN) architecture to analyze medical images. The CNN model is trained using a diverse dataset comprising fundus photographs, OCT scans, and visual field test results. Data preprocessing steps, including image normalization, augmentation, and noise reduction, are applied to improve the quality of the input data. To enhance the model's Выпуск журнала №-26 Часть–6_ Май –2025 93 MODERN EDUCATION AND DEVELOPMENT ability to generalize across different types of eye diseases, transfer learning techniques are employed, leveraging pre-trained models for fine-tuning on our specific dataset. The algorithm is tested and validated using a publicly available dataset, and its performance is evaluated based on accuracy, sensitivity, specificity, and AUC (Area Under the Curve). The AI algorithm demonstrated a high level of performance in detecting signs of diabetic retinopathy, glaucoma, and age-related macular degeneration. For diabetic retinopathy, the algorithm achieved an accuracy of 92%, with sensitivity and specificity values of 90% and 94%, respectively. In the case of glaucoma detection, the algorithm reached an accuracy of 88%, with sensitivity and specificity values of 85% and 89%. For age-related macular degeneration, the model achieved an accuracy of 90%, with sensitivity and specificity of 87% and 92%, respectively. These results su