Scholarlink

Background <p>In a wide range of diseases, it is necessary to utilize multiple data sources to obtain enough data for model training. However, performing centralized pooling of multiple data sources, while protecting each patients’ sensitive data, can require a cumbersome process involving many institutional bodies. Alternatively, federated learning (FL) can be utilized to train models based on data located at multiple sites.</p> Method <p>We propose two methods for training time-to-event prediction models based on distributed data, relying on FL algorithms, for time-to-event prediction models. Both approach incorporates steps to allow prediction of individual-level survival curves, without exposing individual-level event times. For Cox proportional hazards models, the latter is accomplished by using a kernel smoother for the baseline hazard function. The other proposed methodology is based on general parametric likelihood theory for right-censored data. We compared these two methods in four simulation and with one real-world dataset predicting the survival probability in patients with Hodgkin lymphoma (HL).</p> Results <p>The simulations demonstrated that the FL models performed similarly to the non-distributed case in all four experiments, with only slight deviations in predicted survival probabilities compared to the true model. Our findings were similar in the real-world advanced-stage HL example where the FL models were compared to their non-distributed versions, revealing only small deviations in performance.</p> Conclusion <p>The proposed procedures enable training of time-to-event models using data distributed across sites, without direct sharing of individual-level data and event times, while retaining a predictive performance on par with undistributed approaches.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of time to event prediction models using federated learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年5月26日

查看全文
  • Rasmus Rask Kragh Jørgensen,
  • Jonas Faartoft Jensen,
  • Tarec El-Galaly,
  • Martin Bøgsted,
  • Rasmus Froberg Brøndum,
  • Mikkel Runason Simonsen,
  • Lasse Hjort Jakobsen

摘要

Background

In a wide range of diseases, it is necessary to utilize multiple data sources to obtain enough data for model training. However, performing centralized pooling of multiple data sources, while protecting each patients’ sensitive data, can require a cumbersome process involving many institutional bodies. Alternatively, federated learning (FL) can be utilized to train models based on data located at multiple sites.

Method

We propose two methods for training time-to-event prediction models based on distributed data, relying on FL algorithms, for time-to-event prediction models. Both approach incorporates steps to allow prediction of individual-level survival curves, without exposing individual-level event times. For Cox proportional hazards models, the latter is accomplished by using a kernel smoother for the baseline hazard function. The other proposed methodology is based on general parametric likelihood theory for right-censored data. We compared these two methods in four simulation and with one real-world dataset predicting the survival probability in patients with Hodgkin lymphoma (HL).

Results

The simulations demonstrated that the FL models performed similarly to the non-distributed case in all four experiments, with only slight deviations in predicted survival probabilities compared to the true model. Our findings were similar in the real-world advanced-stage HL example where the FL models were compared to their non-distributed versions, revealing only small deviations in performance.

Conclusion

The proposed procedures enable training of time-to-event models using data distributed across sites, without direct sharing of individual-level data and event times, while retaining a predictive performance on par with undistributed approaches.