Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{201375,
        author = {Omkar Rahul Sardesai and Mrs. Tanuja S. Patankar and Mrs. Alpana Borse and Satvik Patil and Rushikesh Raut},
        title = {Federated Learning of Privacy-Preserving Brain Tumor Detection with MRI Image: A Comparative Study},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {3468-3473},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=201375},
        abstract = {One of the most important challenges in clinical diagnosis is the identification of brain tumors from MRI scans, and early discovery greatly improves the patient’s prognosis. In terms of autonomous segmenting and classifying brain tumors, deep learning techniques have recently produced exceptional results. However, collecting sensitive medical data is necessary for traditional model training using a centralized method, which poses serious concerns regarding data security, privacy, and legal issues. Federated learning, which has recently emerged as a possible decentralized learning paradigm, is one of the ways through which many institutions can cooperate using train models without sharing raw patient data. We provide a comparative study of deep learning models, which are based on the work of federated learning to detect brain tumors on MRI images without violating privacy [1].The proposed study evaluates the effectiveness of federated optimization us- ing the Federated Averaging (FedAvg) algorithm in a multi- client setting, simulating distributed healthcare institutions with non-identically distributed (non-IID) data.Lightweight and pre- trained convolutional neural network architectures are utilized, balancing both the computational efficiency with diagnostic accuracy for resource-constrained environments. Experimental analysis reveals competitive performance with federated training while granting data locality and confidentiality. Further, expected trade-offs with respect to centralized learning are noted. The convergence behavior across different communication rounds real-world deployment of federated learning for medical im- age analysis in healthcare [2]. Moreover, challenges including communication efficiency, heterogeneity in systems, and model generalization have been discussed with interpretations from the recent works in federated learning. The obtained results firmly establish FL as a practical and privacy-oriented solution for collaborative brain tumor detection and create a base for further research related to scalable and secure medical imaging.},
        keywords = {Federated Learning, Brain Tumor Detection, Magnetic Resonance Imaging (MRI), Privacy-Preserving Deep Learning},
        month = {May},
        }