Summary of Feddp: Privacy-preserving Method Based on Federated Learning For Histopathology Image Segmentation, by Liangrui Pan et al.
FedDP: Privacy-preserving method based on federated learning for histopathology image segmentation
by Liangrui Pan, Mao Huang, Lian Wang, Pinle Qin, Shaoliang Peng
First submitted to arxiv on: 7 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed federated learning framework, FedDP, enables collaborative training of deep learning models on whole slide images (WSIs) while ensuring the privacy of cancer pathology image data. By employing differential privacy to introduce noise into model updates, the approach minimizes the risk of original data reconstruction through gradient inversion during the training process. The results show that FedDP only slightly impacts model accuracy, with a decrease in Dice, Jaccard, and Acc indices by 0.55%, 0.63%, and 0.42%, respectively. This method facilitates cross-institutional collaboration and knowledge sharing while protecting sensitive data privacy, providing a viable solution for further research and application in the medical field. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Doctors use special stains to look at tissue samples to diagnose cancer. These stainings are important, but it’s hard to share pictures of them because they contain personal information about patients. Researchers came up with an idea to make sure this information stays private while still allowing doctors to learn from each other. They used a special technique called federated learning, which lets different institutions work together without sharing their data. To make sure the data stays safe, they added some noise to the information being shared. This didn’t affect how well the pictures were analyzed, and it allowed doctors to share knowledge while keeping patient information private. |
Keywords
» Artificial intelligence » Deep learning » Federated learning
