Enhancing Medical Image Quality Using Hybrid Deep Neural Networks for Clinical Decision Support

Abstract

This study aims to develop and design an intelligent model based on artificial intelligence and neural network techniques for diagnosing clinical conditions of various diseases, particularly eye diseases, in record time and with high accuracy by improving the quality of medical images. A descriptive analytical methodology was employed, measuring the model's performance using several statistical indicators, such as accuracy, recall rate, and prediction accuracy, to verify its effectiveness and power in analyzing medical images and diagnosing medical conditions. The study also analyzed the model's impact on improving the quality of medical images and its role in helping physicians expedite the diagnostic process and predict potential disease progression, especially in the field of ophthalmology. The results indicated the effectiveness of the proposed model, with an accuracy of 95.8%, a prediction accuracy of 96%, a recall rate of 95.4%, and an F1 score of 94.8%. These indicators reflect the model's efficiency in accurately identifying pathological patterns in medical images. The results also showed a significant improvement in image quality after applying neural network-based processing techniques, with image quality improvement ranging from 28% to 37% compared to the original images. This remarkable improvement can enhance the accuracy of medical diagnoses, reduce the time required for treatment decisions, and increase the clarity of fine details in medical images. This, in turn, supports physicians in diagnosing clinical conditions and taking appropriate treatment measures promptly, especially in cases of eye diseases that demand a high degree of accuracy and speed in diagnosis.