A Master's thesis at the College of Computer Science and Information Technology, University of Basrah, was titled:
(Adversarial Learning for Medical Image Denoising: A GAN-Based Approach)
The thesis, submitted by student Zahraa Makki Hussein, aimed to design and develop an effective model for noise removal from medical images using deep learning techniques, specifically competitive generative adversarial networks (GANs). This model was designed to process various types of medical images and handle multiple types of noise, simulating real-world medical imaging conditions.
The thesis included building a model consisting of a U-Net-based generator, a discriminator, and an additional enhancement network to improve the quality of the resulting images. A set of evaluation metrics was also used to measure the model's efficiency. The thesis concluded that the proposed model achieved outstanding results in improving image quality and preserving diagnostic details, thus enhancing its potential application in clinical medicine.
