During the 27th International Conference On Medical Image Computing And Computer Assisted Intervention (MICCA) the MLCommons® Medical working group presented a  “Federated Learning in Healthcare” tutorial. Previously presented at ISBI 2024, the tutorial taught participants about the Comprehensive Open Federated Ecosystem (COFE). COFE is an open ecosystem of tools and best practices distilled from years of industrial and academic research experience aiming to streamline the research and development of AI/ML models within the clinical setting. 

Figure 1. Updated diagram of latest COFE iteration. HuggingFace Hub was integrated as part of the COFE ecosystem.

During the tutorial the team shared the latest updates in the COFE ecosystem:

  1. Integration of Hugging Face Hub to enable a one-click solution for model dissemination for users, which allows easier research collaboration in the community. Hugging Face Hub also allows for a streamlined approach to a GaNDLF Model Zoo, where users can select appropriate licensing terms (including non-commercial use) for their models..
  2. Full integration with the Medical Open Network for Artificial Intelligence (MONAI) enables developers to reuse engineering and research efforts from different groups while allowing a more cohesive pathway for code reproducibility.
  3. GaNDLF-Synth is a general-purpose synthesis extension of GaNDLF. It provides a vehicle for researchers to train multiple synthesis approaches (autoencoders, GANs, and diffusion) for the same task, thereby allowing them to quantify the efficacy of each approach for their task. Additionally, providing a robust abstraction layer allows algorithmic developers to have a common framework to distribute their methods, allowing easier benchmarking against other approaches.
  4. Privacy-enabled training across various tasks using Opacus. By integrating with Opacus, GaNDLF enables the training of models using differential privacy (DP). DP is one of the most widely used methods to incorporate privacy in trained models, and with GaNDLF, users can now tune their models for various DP settings to achieve the best model utility while training private models.
  5. Single-command model distribution through Hugging Face.
  6. MedPerf’s new web user interface makes evaluating and benchmarking healthcare AI models easier than ever. Designed for seamless interaction, it allows dataset owners and benchmark stakeholders to engage with the platform without requiring technical expertise. This update ensures that anyone involved in the process can contribute to benchmarking effortlessly.

Siddhesh Thakur, a Data Engineer at Indiana University School of Medicine, showcased optimization workflows and their practical application in deploying AI models for whole slide image analysis in resource-constrained environments. Thakur shared key concepts of model optimization and demonstrated how GaNDLF facilitates this process by integrating with Intel’s OpenVINO toolkit. Through online examples, attendees learned how to leverage these tools to create efficient, deployable AI solutions for digital pathology, even in settings with limited computational resources.

Cyril Zakka, a physician and head of ML Research in Health at Hugging Face, demonstrated Hugging Face’s commitment to the democratization of state-of-the-art machine learning and highlighted various open-source tools and offerings for model development, fine-tuning, and deployment across different computing budgets and infrastructures. Dr Zakka outlined the newly developed Health Team’s mission and ongoing projects, highlighting the tight integration within the GaNDLF ecosystem to enable one-click deployment and checkpointing of scientific models, further emphasizing COFE’s goal of reproducibility and scientific accountability.

The community was invited to learn more through a self-guided hands-on session. To learn more about the Medical WG and its work in real-world clinical studies please email [email protected].The hands-on session of this tutorial was made possible through GitHub Codespaces with generous technical support from their engineers and program managers working in close collaboration with the tutorial organizers. The tutorial is available to everyone here.

The team would also like to thank the organizing committee of MICCAI2024 and the conference organizers.

Participants attending the “Federated Learning in Healthcare” at MICCAI 2024 in Marrakesh, Morocco.

Join Us

COFE addresses key challenges that the research community faces in the federated healthcare AI space through open community tools. As an organization, we strive to collect feedback from the broader healthcare research community to help build tools that advance AI/ML for everyone. If you want to participate in the MLCommons Medical working group, please join us here.