Abstract

Federated Learning (FL) enables collaborative, privacy-preserving AI model development on distributed medical data, which is a critical capability for equitable AI in heterogeneous, low-resource settings such as Sub-Saharan Africa (SSA). While FL has been shown feasible for medical imaging in Africa, infrastructure barriers remain poorly characterized before deployment. We conducted a structured infrastructure assessment across 10 SSA clinical and research sites from 10 countries, prior to data collection, using a questionnaire and 1-on-1 interviews with IT personnel, principal investigators, and data managers. Results reveal substantial heterogeneity: imaging formats span DICOM (50%), non-DICOM (40%), and film (10%); clinical data is digital at 40% of sites, paper-based at 50%, and mixed at 10%; storage ranges from dedicated servers to USB drives; and bandwidth spans 1–50 Mbps to over 1 Gbps. Findings are based on a convenience sample of 10 AFRICAI-RI-affiliated sites and may not generalize beyond this network. To conclude, our assessment provides a categorical baseline of infrastructure readiness that can inform planning for federated AI in SSA contexts and similar low-resource settings with comparable infrastructure challenges.

Links to Paper and Supplementary Materials

Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/AFRICAI_023.pdf

SharedIt Link: Not yet available

SpringerLink (DOI): Not yet available

Supplementary Material: Not Submitted

Link to Open Review

Open Review Page: https://openreview.net/forum?id=PyN47fFcsq

BibTex

@InProceedings{BirMah_Infrastructure_MICCAISAT2026,
        author = { Birhanu, Mahlet A. AND Goedhart, Aisha AND Garrucho, Lidia AND Achterberg, Hakim C. AND Lekadir, Karim AND Starmans, Martijn P. A.},
        title = { { Infrastructure Readiness Assessment for Federated Learning in Medical Imaging across African Sites } },
        booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
        year = {2026},
        publisher = {Springer Nature Switzerland},
        volume = {LNCS 17264},
        month = {pending},
        page = {pending}
}


back to top