# Federated Learning in Multi-Center Critical Care Research: A Systematic Case Study using the eICU Database

arXiv: https://arxiv.org/abs/2204.09328 · HTML: https://mlsanity.com/paper/2204.09328/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 17 references that are in the map.
- First posted on arXiv 20 Apr 2022.

## Builds on (most cited first, 7 of 7)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Advances and Open Problems in Federated Learning](https://mlsanity.com/paper/1912.04977/) (2019)
- [Qualitatively characterizing neural network optimization problems](https://mlsanity.com/paper/1412.6544/) (2014)
- [Patient Clustering Improves Efficiency of Federated Machine Learning to predict mortality and hospital stay time using distributed Electronic Medical Records](https://mlsanity.com/paper/1903.09296/) (2019)
- [Benchmarking machine learning models on multi-centre eICU critical care dataset](https://mlsanity.com/paper/1910.00964/) (2019)
- [Federated and Differentially Private Learning for Electronic Health Records](https://mlsanity.com/paper/1911.05861/) (2019)
- [Privacy-Preserving Distributed Deep Learning for Clinical Data](https://mlsanity.com/paper/1812.01484/) (2018)

## Built on by (most cited first, 1 of 1)

- [Client Recruitment for Federated Learning in ICU Length of Stay Prediction](https://mlsanity.com/paper/2304.14663/) (2023)

## Abstract (arXiv)

Federated learning (FL) has been proposed as a method to train a model on different units without exchanging data. This offers great opportunities in the healthcare sector, where large datasets are available but cannot be shared to ensure patient privacy. We systematically investigate the effectiveness of FL on the publicly available eICU dataset for predicting the survival of each ICU stay. We employ Federated Averaging as the main practical algorithm for FL and show how its performance changes by altering three key hyper-parameters, taking into account that clients can significantly vary in size. We find that in many settings, a large number of local training epochs improves the performance while at the same time reducing communication costs. Furthermore, we outline in which settings it is possible to have only a low number of hospitals participating in each federated update round. When many hospitals with low patient counts are involved, the effect of overfitting can be avoided by decreasing the batchsize. This study thus contributes toward identifying suitable settings for running distributed algorithms such as FL on clinical datasets.