ScholarDeck

AI / informaticsJournal of clinical monitoring and computing · 2026

Continuous imputation of blood gas and metabolic panel laboratory values in the intensive care unit using machine learning.

Behrooz Mamandipoor, Martin Krause, Pragnya Korti … Rodney A Gabriel

Pulmonology & Critical CareAI / informatics

Machine learning imputation yielded higher accuracy predicting need for renal replacement therapy

Continuous imputation of blood gas and metabolic panel laboratory values in the intensive care unit using machine learning.

Behrooz Mamandipoor et al. · Journal of clinical monitoring and computing · 2026

Methods

We evaluated its impact on prediction of the need for renal replacement therapy (RRT).

Results

computer's ability to correctly flag patients needing dialysis was very high

ML imputation

0.951AUROC

Previous-value

0.923AUROC

More results

We compared the performance of the deep learning models using laboratory variables imputed hourly by our ML estimators versus models using laboratory variables imputed by a previous-value baseline.

Accuracy of laboratory imputations were compared using mean absolute error (MAE) and root mean squared error (RMSE).

More results

We compared performance using area under the receiver operating characteristics curve (AUROC) and area under the precision-recall curve (AUPRC).

XGBoost achieved an average error reduction of 33% in MAE and 32% in RMSE across ABG variables, and 19% and 21% across BMP variables, respectively.

“

Conclusion

Our real-time imputation models led to significant error reduction of most ABG and BMP targets, improving deep learning model performance for the need of RRT.

Post

Continuous imputation of blood gas and metabolic panel laboratory values in the intensive care unit using machine learning.

Machine learning imputation yielded higher accuracy predicting need for renal replacement therapy

Related papers