The Institute of Informatics for Graduate Studies at the University of Information Technology and Communications discussed a master’s thesis submitted by student Noor Sabah Hassan in the field of Computer Science, entitled:

“Medical Image Classification for Multi Healthcare Centers Using Federated Learning”

The thesis aimed to study and evaluate the efficiency of federated learning techniques in classifying medical images within multi-center healthcare environments. This was achieved through a comparison between two main architectural frameworks: Centralized Federated Learning (CFL) and Decentralized Federated Learning (DFL). The study also analyzed the performance of these models in terms of accuracy, convergence speed, stability, and generalizability, based on an Iraqi dataset on gallbladder diseases.

The researcher employed Intended Intended and Non-Intended Partitioning (IID) methods to simulate levels of homogeneity and statistical variance among three and six collaborating medical centers. This approach contributes to enhancing the efficiency of intelligent models while maintaining the privacy of medical data.

At the conclusion of the discussion, the student was awarded a Master’s degree with a grade of “Very Good” after the scientific committee commended the study’s findings and its contribution to developing artificial intelligence applications in the healthcare sector.