<![CDATA[Advances in artificial intelligence technology, particularly deep learning, have been widely utilized in medical image processing to support early disease detection. Previous studies have shown that Convolutional Neural Networks (CNN) are capable of achieving high levels of accuracy in medical image classification, but differences in architecture and training methods result in varying performance. This study aims to analyze and compare the performance of deep learning algorithms in early disease detection using medical images, using previous research as a benchmark. Transfer learning-based CNN models, namely VGG16 and ResNet50, were used and evaluated using a labeled medical image dataset. Performance evaluation was conducted using accuracy, precision, recall, and F1-score metrics, and the results were compared with accuracy achieved in previous studies using similar approaches. The analysis showed that the ResNet50 model achieved an accuracy of up to 95.3%, comparable to or better than several previous studies. These findings confirm that the choice of CNN architecture and training strategy significantly influences the performance of medical image-based early disease detection systems.]]>
Copyrights © 2026
