Interpretable Machine Learning Model Using Oxygenation and Respiratory Variability to Predict Hemorrhagic…

AI / informaticsComputer methods and programs in biomedicine· Jul 2026

Interpretable Machine Learning Model Using Oxygenation and Respiratory Variability to Predict Hemorrhagic Stroke Mortality: A Multicenter Validation Study.

Jing Feng, Hongyu Zhang, Jianheng Gu … Baojian Wei

Prediction Models & Risk ScoresAI / informaticsPrognosisQ1IF 6.7

Oxygenation-based machine learning model showed acceptable external accuracy predicting hemorrhagic stroke ICU mortality.

Interpretable Machine Learning Model Using Oxygenation and Respiratory Variability to Predict Hemorrhagic Stroke Mortality: A Multicenter Validation Study.

Jing Feng … Baojian Wei

Computer methods and programs in biomedicine · Jul 2026 · Q1 in Health Informatics

Background

Hemorrhagic stroke has high in-hospital mortality, but early risk prediction remains suboptimal.

Purpose

To address this gap, we aimed to evaluate the prognostic value of oxygenation and respiratory variability and incorporate it into an interpretable machine learning framework for mortality prediction in hemorrhagic stroke.

Methods

From MIMIC-IV were used for model development and internal validation (n = 2,262), and patients from the eICU Collaborative Research Database were used for external validation (n = 2,076).

n = 2,262

Results

correctly told apart survivors from non-survivors most of the time, though not as reliably as the internal test

Internal validation

0.911AUROC

External validation

0.792AUROC

More results

In the external cohort, the model maintained a high negative predictive value but had a low positive predictive value, indicating limited precision for identifying high-risk patients at the applied threshold.

More results

SHAP analysis identified HR-min, Glucose-mean, Sodium-mean, and several oxygenation and respiratory variability features, including TPE-RR, Slope-PaO₂, Mean-SpO₂, and SD-SpO₂, as important contributors to mortality prediction.

“

Conclusion · 1 of 3

We developed and externally validated a machine learning model incorporating oxygenation and respiratory variability for mortality risk estimation in patients with hemorrhagic stroke.

Conclusion · 2 of 3

The model showed strong internal performance and acceptable external discrimination after a 72-hour ICU observation window.

Conclusion · 3 of 3

However, the low external positive predictive value and cross-cohort differences in event prevalence and treatment patterns limit immediate clinical application.

Interpretable Machine Learning Model Using Oxygenation and Respiratory Variability to Predict Hemorrhagic Stroke Mortality: A Multicenter Validation Study.

Interpretable Machine Learning Model Using Oxygenation and Respiratory Variability to Predict Hemorrhagic Stroke Mortality: A Multicenter Validation Study.

Related papers