Towards an Explainable Model for Sepsis Detection Based on Sensitivity Analysis - 06/07/21

Doi : 10.1016/j.irbm.2021.05.006

M. Chen ^a, A. Hernández ^b,^{^⁎}
^a The Laboratory of Medical Imaging and Artificial Intelligence, School of Biomedical Science and Medical Engineering, Southeast University, Nanjing, China
^b Université de Rennes, Inserm, LTSI - UMR 1099, Rennes, France

^⁎Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Tuesday 06 July 2021
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Abstract

Objectives

Sepsis is a life-threatening condition which is responsible for a high proportion of intra-hospital deaths and related healthcare costs each year. Early detection and treatment of sepsis episodes is critical, since an early treatment may highly improve prognosis. This study proposed an original method to increase the interpretability of a set of machine learning models for the early detection of sepsis onset.

Material and methods

Open data from the electronic medical records of 40,336 patients monitored in intensive care units (ICU), provided by the PhysioNet/Computing in Cardiology Challenge 2019 is used in this paper. We proposed a method including data preprocessing, feature engineering, model construction and tuning, as well as an original interpretability analysis method for the final stage.

Results

A total of 24 models were developed and analyzed. The best model, based on 142 features achieved a 0.4274 utility score. The best compact model integrates only 20 selected features, and provided a utility score of 0.3862. Meanwhile, the proposed sensitivity analysis method allows for the identification of the most relevant markers to early detect the onset of sepsis, as well as their interdependence and relative importance on the final decision.

Conclusion

A set of optimized machine-learning models were proposed for predicting sepsis early in a real-time way with high performance, and interpretable information including the most significant biomarkers were analyzed through novel interpretability method.

Le texte complet de cet article est disponible en PDF.

Graphical abstract

Le texte complet de cet article est disponible en PDF.

Highlights

•	Optimized machine learning pipelines for Sepsis early prediction are proposed.
•	The models are fitted on the 2019 PhysioNet/Computing in Cardiology Challenge dataset.
•	The models achieved comparable performance to the best methods of the challenge.
•	A new method for increasing the interpretability of machine learning models is proposed.
•	Interpretable information was extracted from the proposed model.