Federated Learning for 2D Medical Image Registration: A Comprehensive Survey

This survey has examined federated learning applied to 2D medical image registration. FL fits this problem naturally: registration losses are unsupervised, so every participating institution contributes to training regardless of whether its data is annotated, while the federated protocol keeps raw images local. We covered the full technical stack—classical and deep-learning registration, FL aggregation and non-IID challenges, privacy mechanisms, and domain adaptation—and proposed an evaluation protocol that accounts for cross-site generalisation, personalisation, and the privacy–utility trade-off. The intersection of FL and 2D registration is still early-stage. Adaptive regularisation, registration-specific privacy mechanisms, and genuine multi-site benchmarks are the most immediate gaps. We hope this survey is a useful starting point for researchers entering the area.

[1]     a review,” Comput. Methods Biomech. Biomed. Eng., vol. 17, no. 2, pp. 73–93, 2014. [2]     G. Balakrishnan, A. Zhao, M. R. Sabuncu, J. Guttag, and A. V. Dalca, “VoxelMorph: a learning framework for deformable medical image registration,” IEEE Trans. Med. Imag., vol. 38, no. 8, pp. 1788–1800, Aug. 2019. [3]     M. Jaderberg, K. Simonyan, A. Zisserman, and K. Kavukcuoglu, “Spa-tial transformer networks,” in Proc. Adv. Neural Inf. Process. Syst. (NeurIPS), vol. 28, 2015, pp. 2017–2025. [4]     H. B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. Y. Arcas, “Communication-efficient learning of deep networks from decentralized data,” in Proc. Int. Conf. Artif. Intell. Stat. (AISTATS), 2017, pp. 1273–1282. [5]     T. Li, A. K. Sahu, M. Zaheer, M. Sanjabi, A. Smola, and V. Smith, “Federated optimization in heterogeneous networks,” in Proc. Mach. Learn. Syst. (MLSys), 2020. [6]     B. K. P. Horn and B. G. Schunck, “Determining optical flow,” Artif. Intell., vol. 17, no. 1–3, pp. 185–203, 1981. [7]     J. P. Thirion, “Image matching as a diffusion process: an analogy with Maxwell’s demons,” Med. Image Anal., vol. 2, no. 3, pp. 243–260, 1998. [8]     D. Rueckert, L. I. Sonoda, C. Hayes, D. L. G. Hill, M. O. Leach, and D. J. Hawkes, “Nonrigid registration using free-form deformations: application to breast MR images,” IEEE Trans. Med. Imag., vol. 18, no. 8, pp. 712–721, Aug. 1999. [9]     O. Ronneberger, P. Fischer, and T. Brox, “U-Net: convolutional networks for biomedical image segmentation,” in Proc. Int. Conf. Med. Image Comput. Comput.-Assist. Interv. (MICCAI), 2015, pp. 234–241. [10]    C. Dwork, F. McSherry, K. Nissim, and A. Smith, “Calibrating noise to sensitivity in private data analysis,” in Proc. Theory Cryptogr. Conf. (TCC), 2006, pp. 265–284.