Performance comparison between federated and centralized learning with a deep learning model on Hoechst stained images

Medical data is not fully exploited by Machine Learning (ML) techniques because the privacy concerns restrict the sharing of sensitive information and consequently the use of centralized ML schemes. Usually, ML models trained on local data are failing to reach their full potential owing to low statistical power.
Federated Learning (FL) solves critical issues in the healthcare domain such as data privacy and enables multiple contributors to build a common and robust ML model by sharing local learning parameters without sharing data. FL approaches are mainly evaluated in the literature using benchmarks and the trade-off between accuracy and privacy still has to be more studied in realistic clinical contexts.
In this work, we evaluate this trade-off for a CD3/CD8 cells labeling model from Hoechst stained images. Wölflein et al. developed a deep learning GAN model that labels CD3 and CD8 cells from kidney cancer tissue slides stained with Hoechst. The GAN model was trained on 475,000 patches (256x256 pixels) from 8 whole slide images.
We modified the training to simulate a FL approach by distributing the learning across several clients and aggregating the parameters to create the overall model. We present the performance comparison between FL and centralized learning.