WRSE - a non-parametric weighted-resolution ensemble for predicting individual survival distributions in the ICU

Jonathan Heitz, Joanna Ficek, Martin Faltys, Tobias M. Merz, Gunnar Rätsch, Matthias Hüser
Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021, PMLR 146:54-69, 2021.

Abstract

Dynamic assessment of mortality risk in the intensive care unit (ICU) can be used to stratify patients, inform about treatment effectiveness or serve as part of early-warning systems. Static risk scores, such as APACHE or SAPS, have been supplemented with data-driven approaches that track dynamic mortality risk over time. Recent works have focused on enhancing the information delivered to clinicians even further by producing full survival distributions instead of point predictions or fixed horizon risks. In this work, we propose a non-parametric ensemble model, Weighted Resolution Survival Ensemble (WRSE), tailored to estimate such dynamic individual survival distributions. Inspired by the simplicity and robustness of ensemble methods, the proposed approach combines a set of binary classifiers spaced according to a decay function reflecting the relevance of short-term predictions. Models and baselines are evaluated under weighted calibration and discrimination metrics for individual survival distributions, which closely reflect the utility of a model in ICU practice. We show competitive results with state-of-the-art probabilistic models, while greatly reducing training time by factors of 2-9x.

Cite this Paper


BibTeX
@InProceedings{pmlr-v146-heitz21a, title = {WRSE - a non-parametric weighted-resolution ensemble for predicting individual survival distributions in the ICU}, author = {Heitz, Jonathan and Ficek, Joanna and Faltys, Martin and Merz, Tobias M. and R{\"a}tsch, Gunnar and H{\"u}ser, Matthias}, booktitle = {Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021}, pages = {54--69}, year = {2021}, editor = {Greiner, Russell and Kumar, Neeraj and Gerds, Thomas Alexander and van der Schaar, Mihaela}, volume = {146}, series = {Proceedings of Machine Learning Research}, month = {22--24 Mar}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v146/heitz21a/heitz21a.pdf}, url = {http://proceedings.mlr.press/v146/heitz21a.html}, abstract = {Dynamic assessment of mortality risk in the intensive care unit (ICU) can be used to stratify patients, inform about treatment effectiveness or serve as part of early-warning systems. Static risk scores, such as APACHE or SAPS, have been supplemented with data-driven approaches that track dynamic mortality risk over time. Recent works have focused on enhancing the information delivered to clinicians even further by producing full survival distributions instead of point predictions or fixed horizon risks. In this work, we propose a non-parametric ensemble model, Weighted Resolution Survival Ensemble (WRSE), tailored to estimate such dynamic individual survival distributions. Inspired by the simplicity and robustness of ensemble methods, the proposed approach combines a set of binary classifiers spaced according to a decay function reflecting the relevance of short-term predictions. Models and baselines are evaluated under weighted calibration and discrimination metrics for individual survival distributions, which closely reflect the utility of a model in ICU practice. We show competitive results with state-of-the-art probabilistic models, while greatly reducing training time by factors of 2-9x.} }
Endnote
%0 Conference Paper %T WRSE - a non-parametric weighted-resolution ensemble for predicting individual survival distributions in the ICU %A Jonathan Heitz %A Joanna Ficek %A Martin Faltys %A Tobias M. Merz %A Gunnar Rätsch %A Matthias Hüser %B Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021 %C Proceedings of Machine Learning Research %D 2021 %E Russell Greiner %E Neeraj Kumar %E Thomas Alexander Gerds %E Mihaela van der Schaar %F pmlr-v146-heitz21a %I PMLR %P 54--69 %U http://proceedings.mlr.press/v146/heitz21a.html %V 146 %X Dynamic assessment of mortality risk in the intensive care unit (ICU) can be used to stratify patients, inform about treatment effectiveness or serve as part of early-warning systems. Static risk scores, such as APACHE or SAPS, have been supplemented with data-driven approaches that track dynamic mortality risk over time. Recent works have focused on enhancing the information delivered to clinicians even further by producing full survival distributions instead of point predictions or fixed horizon risks. In this work, we propose a non-parametric ensemble model, Weighted Resolution Survival Ensemble (WRSE), tailored to estimate such dynamic individual survival distributions. Inspired by the simplicity and robustness of ensemble methods, the proposed approach combines a set of binary classifiers spaced according to a decay function reflecting the relevance of short-term predictions. Models and baselines are evaluated under weighted calibration and discrimination metrics for individual survival distributions, which closely reflect the utility of a model in ICU practice. We show competitive results with state-of-the-art probabilistic models, while greatly reducing training time by factors of 2-9x.
APA
Heitz, J., Ficek, J., Faltys, M., Merz, T.M., Rätsch, G. & Hüser, M.. (2021). WRSE - a non-parametric weighted-resolution ensemble for predicting individual survival distributions in the ICU. Proceedings of AAAI Spring Symposium on Survival Prediction - Algorithms, Challenges, and Applications 2021, in Proceedings of Machine Learning Research 146:54-69 Available from http://proceedings.mlr.press/v146/heitz21a.html.

Related Material