A Federated Cox Model with Non-Proportional Hazards
This addresses the need for privacy-preserving survival models in healthcare, though it is incremental as it builds on existing federated learning and Cox model extensions.
The paper tackled the problem of applying neural networks to survival analysis in healthcare where data is decentralized and the proportional hazards assumption may not hold, by presenting a federated Cox model that accommodates time-varying covariate effects without explicit specification, and demonstrated it performs as well as a standard model on clinical datasets.
Recent research has shown the potential for neural networks to improve upon classical survival models such as the Cox model, which is widely used in clinical practice. Neural networks, however, typically rely on data that are centrally available, whereas healthcare data are frequently held in secure silos. We present a federated Cox model that accommodates this data setting and also relaxes the proportional hazards assumption, allowing time-varying covariate effects. In this latter respect, our model does not require explicit specification of the time-varying effects, reducing upfront organisational costs compared to previous works. We experiment with publicly available clinical datasets and demonstrate that the federated model is able to perform as well as a standard model.