RMS: A ML-based system for ICU Respiratory Monitoring and Resource Planning

Hüser, Matthias; Lyu, Xinrui; Faltys, Martin; Pace, Alizée; Berger, David; Hoche, Marine; Hyland, Stephanie L; Yèche, Hugo; Burger, Manuel; Merz, Tobias M; Rätsch, Gunnar

doi:10.1101/2024.01.23.24301516

Cited by 3 publications

(8 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The HiRID-II data set (Hüser et al, 2024) was used to test performance in the Swiss health system. The dataset was previously k-anonymized with respect to observable attributes including age, gender, weight and height and absolute date information was removed (Hüser et al, 2024). The MIMIC-IV data set (Goldberger et al, 2000; Johnson et al, 2023) was used to assess performance in the US health system.…”

Section: Methodsmentioning

confidence: 99%

“…For an exhaustive description of these features we refer to Hyland et al, 2020 and Hüser et al, 2024.…”

Section: Methodsmentioning

confidence: 99%

“…The intersection of these 5 sets yielded 23 variables. Then greedy forward selection on this set guided by AUPRC on the validation set was performed in the 5 splits, and the first k variables were included such that the model’s was within 1 % of the optimal performance, as done in related works (Hyland et al, 2020; Hüser et al, 2024). The variable sets found in the 5 splits were then intersected, yielding 15 variables.…”

Section: Methodsmentioning

confidence: 99%

“…For HiRID-II, we used the provided split into development and test set (Hüser et al, 2024). The latter comprised the latest admissions of the HiRID-II dataset from June 2018-June 2019, to evaluate whether models generalize well into the future.…”

Section: Data Splitting Setupmentioning

confidence: 99%

“…For similar prediction tasks, machine learning approaches have been shown to thrive in a data-rich environment such as the ICU. Examples include early circulatory (Hy-land et al, 2020) and respiratory failure prediction (Hüser et al, 2024). A number of challenges need to be overcome for such a predictor to effectively support the workflow of ICU clinicians.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

An Empirical Study on KDIGO-Defined Acute Kidney Injury Prediction in the Intensive Care Unit

Lyu,

Fan,

Hüser

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Motivation: Acute kidney injury (AKI) is a syndrome that affects a large fraction of all critically ill patients, and early diagnosis to receive adequate treatment is as imperative as it is challenging to make early. Consequently, machine learning approaches have been developed to predict AKI ahead of time. However, the prevalence of AKI is often underestimated in state-of-the-art approaches, as they rely on an AKI event annotation solely based on creatinine, ignoring urine output. Methods: We construct and evaluate early warning systems for AKI in a multi-disciplinary ICU setting, using the complete KDIGO definition of AKI. We propose several variants of gradient-boosted decision trees (GBDT)-based models, including a novel time-stacking based approach. A state-of-the-art LSTM-based model previously proposed for AKI prediction is used as a comparison, which was not specifically evaluated in ICU settings yet. Results: We find that optimal performance is achieved by using GBDT with the time-based stacking technique (AUPRC=65.7%, compared with the LSTM-based model's AUPRC=62.6%), which is motivated by the high relevance of time since ICU admission for this task. Both models show mildly reduced performance in the limited training data setting, perform fairly across different subcohorts, and exhibit no issues in gender transfer. Conclusion: Following the official KDIGO definition substantially increases the number of annotated AKI events. In our study GBDTs outperform LSTM models for AKI prediction. Generally, we find that both model types are robust in a variety of challenging settings arising for ICU data.

show abstract

Section: Methodsmentioning

confidence: 99%

“…For an exhaustive description of these features we refer to Hyland et al, 2020 and Hüser et al, 2024.…”

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Data Splitting Setupmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

An Empirical Study on KDIGO-Defined Acute Kidney Injury Prediction in the Intensive Care Unit

Lyu,

Fan,

Hüser

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

An empirical study on KDIGO-defined acute kidney injury prediction in the intensive care unit

Lyu,

Fan,

Hüser

et al. 2024

Bioinformatics

View full text Add to dashboard Cite

Motivation Acute kidney injury (AKI) is a syndrome that affects a large fraction of all critically ill patients, and early diagnosis to receive adequate treatment is as imperative as it is challenging to make early. Consequently, machine learning approaches have been developed to predict AKI ahead of time. However, the prevalence of AKI is often underestimated in state-of-the-art approaches, as they rely on an AKI event annotation solely based on creatinine, ignoring urine output. We construct and evaluate early warning systems for AKI in a multi-disciplinary ICU setting, using the complete KDIGO definition of AKI. We propose several variants of gradient-boosted decision tree (GBDT)-based models, including a novel time-stacking based approach. A state-of-the-art LSTM-based model previously proposed for AKI prediction is used as a comparison, which was not specifically evaluated in ICU settings yet. Results We find that optimal performance is achieved by using GBDT with the time-based stacking technique (AUPRC = 65.7%, compared with the LSTM-based model’s AUPRC = 62.6%), which is motivated by the high relevance of time since ICU admission for this task. Both models show mildly reduced performance in the limited training data setting, perform fairly across different subcohorts, and exhibit no issues in gender transfer. Following the official KDIGO definition substantially increases the number of annotated AKI events. In our study GBDTs outperform LSTM models for AKI prediction. Generally, we find that both model types are robust in a variety of challenging settings arising for ICU data. Availability and implementation The code to reproduce the findings of our manuscript can be found at: https://github.com/ratschlab/AKI-EWS

show abstract

FAMEWS: a Fairness Auditing tool for Medical Early-Warning Systems

Hoche,

Mineeva,

Burger

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Machine learning applications hold promise to aid clinicians in a wide range of clinical tasks, from diagnosis to prognosis, treatment, and patient monitoring. These potential applications are accompanied by a surge of ethical concerns surrounding the use of Machine Learning (ML) models in healthcare, especially regarding fairness and non-discrimination. While there is an increasing number of regulatory policies to ensure the ethical and safe integration of such systems, the translation from policies to practices remains an open challenge. Algorithmic frameworks, aiming to bridge this gap, should be tailored to the application to enable the translation from fundamental human-right principles into accurate statistical analysis, capturing the inherent complexity and risks associated with the system. In this work, we propose a set of fairness impartial checks especially adapted to ML early-warning systems in the medical context, comprising on top of standard fairness metrics, an analysis of clinical outcomes, and a screening of potential sources of bias in the pipeline. Our analysis is further fortified by the inclusion of event-based and prevalence-corrected metrics, as well as statistical tests to measure biases. Additionally, we emphasize the importance of considering subgroups beyond the conventional demographic attributes. Finally, to facilitate operationalization, we present an open-source tool FAMEWS to generate comprehensive fairness reports. These reports address the diverse needs and interests of the stakeholders involved in integrating ML into medical practice. The use of FAMEWS has the potential to reveal critical insights that might otherwise remain obscured. This can lead to improved model design, which in turn may translate into enhanced health outcomes.

show abstract

RMS: A ML-based system for ICU Respiratory Monitoring and Resource Planning

Cited by 3 publications

References 67 publications

An Empirical Study on KDIGO-Defined Acute Kidney Injury Prediction in the Intensive Care Unit

An Empirical Study on KDIGO-Defined Acute Kidney Injury Prediction in the Intensive Care Unit

An empirical study on KDIGO-defined acute kidney injury prediction in the intensive care unit

FAMEWS: a Fairness Auditing tool for Medical Early-Warning Systems

Contact Info

Product

Resources

About