Scholarlink

Background <p>Patient-reported survey data are used to train prognostic models aimed at improving healthcare. However, such data are typically available multi-centric and, for privacy reasons, cannot easily be centralized in one data repository. Models trained locally are less accurate, robust, and generalizable. We aim to investigate the applicability of privacy-preserving federated machine learning techniques for prognostic model building on health survey data, where local data never leaves the legally safe harbors of the medical centers.</p> Methods <p>We used centralized, local, and federated learning techniques on two healthcare datasets (<i>GLA: D</i><sup>®</sup>data from the five health regions of Denmark and international SHARE data of 27 countries) to predict two different health outcomes. We compared linear regression, random forest regression, and random forest classification models trained on local data with those trained on the entire data in a centralized and in a federated fashion.</p> Results <p>In GLA: D<sup>®</sup> data, federated linear regression (R<sup><i>2</i></sup> 0.34, RMSE 18.2) and federated random forest regression (R<sup><i>2</i></sup> 0.34, RMSE 18.3) models outperform their local counterparts (i.e., R<sup><i>2</i></sup> 0.32, RMSE 18.6, R<sup><i>2</i></sup> 0.30, RMSE 18.8) with statistical significance. We also found that centralized models (R<sup><i>2</i></sup> 0.34, RMSE 18.2, R<sup><i>2</i></sup> 0.32, RMSE 18.5, respectively) did not perform significantly better than the federated models. In SHARE, the federated model (AC 0.78, AUROC: 0.71) and centralized model (AC 0.84, AUROC: 0.66) perform significantly better than the local models (AC: 0.74, AUROC: 0.69).</p> Conclusion <p>Federated learning enables the training of prognostic models from multi-center surveys without compromising privacy and with only minimal or no compromise regarding model performance.</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

Privacy-preserving federated prediction of health outcomes using multi-center survey data

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月4日

查看全文
  • Supratim Das,
  • Mahdie Rafiei,
  • Paula T. Kammer,
  • Søren T. Skou,
  • Dorte T. Grønne,
  • Ewa M. Roos,
  • André Hajek,
  • Hans-Helmut König,
  • Md Shihab Ullah,
  • Niklas Probul,
  • Jan Baumbach,
  • Linda Baumbach

摘要

Background

Patient-reported survey data are used to train prognostic models aimed at improving healthcare. However, such data are typically available multi-centric and, for privacy reasons, cannot easily be centralized in one data repository. Models trained locally are less accurate, robust, and generalizable. We aim to investigate the applicability of privacy-preserving federated machine learning techniques for prognostic model building on health survey data, where local data never leaves the legally safe harbors of the medical centers.

Methods

We used centralized, local, and federated learning techniques on two healthcare datasets (GLA: D®data from the five health regions of Denmark and international SHARE data of 27 countries) to predict two different health outcomes. We compared linear regression, random forest regression, and random forest classification models trained on local data with those trained on the entire data in a centralized and in a federated fashion.

Results

In GLA: D® data, federated linear regression (R2 0.34, RMSE 18.2) and federated random forest regression (R2 0.34, RMSE 18.3) models outperform their local counterparts (i.e., R2 0.32, RMSE 18.6, R2 0.30, RMSE 18.8) with statistical significance. We also found that centralized models (R2 0.34, RMSE 18.2, R2 0.32, RMSE 18.5, respectively) did not perform significantly better than the federated models. In SHARE, the federated model (AC 0.78, AUROC: 0.71) and centralized model (AC 0.84, AUROC: 0.66) perform significantly better than the local models (AC: 0.74, AUROC: 0.69).

Conclusion

Federated learning enables the training of prognostic models from multi-center surveys without compromising privacy and with only minimal or no compromise regarding model performance.