A hybrid modeling framework for generalizable and interpretable predictions of ICU mortality across multiple hospitals

Scientific Reports, vol. 14, pp. 5725

Abstract

The development of reliable mortality risk stratification models is an active research area in computational healthcare. Mortality risk stratification provides a standard to assist physicians in evaluating a patient's condition or prognosis objectively. Particular interest lies in methods that are transparent to clinical interpretation and that retain predictive power once validated across diverse datasets they were not trained on. This study addresses the challenge of consolidating numerous ICD codes for predictive modeling of ICU mortality, employing a hybrid modeling approach that integrates mechanistic, clinical knowledge with mathematical and machine learning models . A tree-structured network connecting independent modules that carry clinical meaning is implemented for interpretability. Our training strategy utilizes graph-theoretic methods for data analysis, aiming to identify the functions of individual black-box modules within the tree-structured network by harnessing solutions from specific max-cut problems. The trained model is then validated on external datasets from different hospitals, demonstrating successful generalization capabilities, particularly in binary-feature datasets where label assessment involves extrapolation.

Authors 7

1. Moein E. Samadi corresponding Aachen Institute for Computational Biomedicine

   RWTH Aachen University

   Affiliation as printed

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany. moein.samadi@rwth-aachen.de

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany

2. Jorge Guzman-Maldonado Aachen Institute for Computational Biomedicine

   RWTH Aachen University

   Affiliation as printed

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany

3. Kateryna Nikulina Aachen Institute for Computational Biomedicine

   RWTH Aachen University

   Affiliation as printed

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany

4. Hedieh Mirzaieazar Aachen Institute for Computational Biomedicine

   RWTH Aachen University

   Affiliation as printed

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany

5. Konstantin Sharafutdinov Aachen Institute for Computational Biomedicine

   RWTH Aachen University

   Affiliation as printed

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany

6. Sebastian Johannes Fritsch Aachen Department of Intensive Care Medicine

   Forschungszentrum Jülich · Universitätsklinikum Aachen · Jülich Supercomputing Centre · RWTH Aachen University

   Affiliation as printed

   Center for Advanced Simulation and Analytics (CASA), Forschungszentrum Jülich, Jülich, Germany

   Department of Intensive Care Medicine, University Hospital RWTH Aachen, Aachen, Germany

   Jülich Supercomputing Centre, Forschungszentrum Jülich, Jülich, Germany

7. Andreas Schuppert Aachen Institute for Computational Biomedicine

   RWTH Aachen University

   Affiliation as printed

   Institute for Computational Biomedicine, RWTH Aachen University, Aachen, Germany