Skip to content

[MICCAI2024] "FedFMS: Exploring Federated Foundation Models for Medical Image Segmentation". A framework for fine-tuning SAM (Segment Anything) in the federated learning paradigm for medical image segmentation.

Notifications You must be signed in to change notification settings

LIU-YUXI/FedFMS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

32 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FedFMS

😆 Introdcution

Develop federated foundation models for medical image segmentation.

(1) A framework for fine-tuning all the parameters of SAM (Segment Anything Model) in the federated learning paradigm for medical image segmentation, called FedSAM.

(2) A framework for fine-tuning MSA (Medical SAM Adapter) in the federated learning paradigm for medical image segmentation, called FedMSA.

Welcome to read our paper, which provides detailed descriptions of the methods and experimental results: FedFMS: Exploring Federated Foundation Models for Medical Image Segmentation

FedFMS: Exploring Federated Foundation Models for Medical Image Segmentation
Yuxi Liu, Guibo Luo*, Yuesheng Zhu* MICCAI2024

🌟 Citation

If you find this work is helpful to your research, please consider citing our paper:

@article{liu2024fedfms,
  title={FedFMS: Exploring Federated Foundation Models for Medical Image Segmentation},
  author={Liu, Yuxi and Luo, Guibo and Zhu, Yuesheng},
  journal={arXiv preprint arXiv:2403.05408},
  year={2024}
}

Thanks for your interest in our work!

📝 Requirements

  • Python==3.8.16

  • torch==2.0.0

  • torchvision==0.15.0

  • numpy==1.24.3

  • opencv_python==4.7.0.72

See the detail requirements in requirements.txt

🚀 Model Training and Test

run FedSAM

run ./run_ft/run_fed_sam_FeTS_ft.sh for federated learning in FeTS2022 Dataset (Brain Tumor).

run ./run_ft/run_fed_sam_fundus_ft.sh for federated learning in Fundus Dataset.

run ./run_ft/run_fed_sam_nuclei_ft.sh for federated learning in Nuclei Dataset.

run ./run_ft/run_fed_sam_prostate_ft.sh for federated learning in Prostate Cancer Dataset.

run ./run_ft/run_fed_sam_ctlung_ft.sh for federated learning in Lung Dataset.

run FedMSA

run ./run_fed_sam_FeTS.sh for federated learning in FeTS2022 Dataset (Brain Tumor).

run ./run_fed_sam_fundus.sh for federated learning in Fundus Dataset.

run ./run_fed_sam_nuclei.sh for federated learning in Nuclei Dataset.

run ./run_fed_sam_prostate.sh for federated learning in Prostate Cancer Dataset.

run ./run_fed_sam_ctlung.sh for federated learning in Lung Dataset.

📚 Data sources

FeTS2022 Dataset

Federated Tumor Segmentation Challenge | FeTS Challenge 2021 (fets-ai.github.io)

Fundus Dataset

REFUGE Challenge Dataset | Papers With Code

Glaucoma Fundus Imaging Datasets | Kaggle

Drishti-GS - RETINA DATASET FOR ONH SEGMENTATION (kaggle.com)

Nuclei Dataset

Prostate cANcer graDe Assessment (PANDA) Challenge | Kaggle

MoNuSAC-2020 - Home (biomedicalimaging.org)

TNBC_dataset | Kaggle

MonuSeg-2018 | Kaggle

Prostate Cancer Dataset

Evaluation of prostate segmentation algorithms for MRI: The PROMISE12 challenge - ScienceDirect

Initiative for Collaborative Computer Vision Benchmarking (i2cvb.github.io)

NCI-ISBI 2013 Challenge - Automated Segmentation of Prostate Structures - The Cancer Imaging Archive (TCIA) Public Access - Cancer Imaging Archive Wiki

Lung Dataset

CT Lung & Heart & Trachea segmentation | Kaggle

About

[MICCAI2024] "FedFMS: Exploring Federated Foundation Models for Medical Image Segmentation". A framework for fine-tuning SAM (Segment Anything) in the federated learning paradigm for medical image segmentation.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published