Post

New research · Pulmonology & Critical Care
NPJ digital medicine · 2d
Cohort studyNPJ digital medicine · 2026

Expert Augmented Prediction of Circulatory and Respiratory Instability from High Resolution Vital Signs.

Luhao Wang, Bin Gu, Yao Nie … Xiangdong Guan
Read paper
Pulmonology & Critical CareCohort study

Machine learning models accurately predict circulatory and respiratory instability in ICUs.

Expert Augmented Prediction of Circulatory and Respiratory Instability from High Resolution Vital Signs.

Luhao Wang et al. · NPJ digital medicine · 2026
Methods

Machine‑learning models were trained on 627,958 h of continuous vital‑sign data from 1702 ICU patients at the First Affiliated Hospital of Sun Yat‑sen University and externally validated in the MIMIC‑III cohort.

n = 1702 ICU patients
Results
The models were highly accurate at predicting circulatory and respiratory instability
n = 1702 ICU patients
More results

Leveraging routinely collected high-frequency vital-sign monitoring, we developed an interpretable, expert-augmented early warning system based on 1-second-resolution heart rate, blood pressure, respiratory rate, and oxygen saturation data.

More results

Increasing temporal resolution improved predictive accuracy, with trend-based features contributing most strongly to model predictions.

“
Conclusion

EAEWS generated accurate, low-frequency alerts with transparent explanations aligned with bedside monitoring, and may provide a scalable framework for real-time CRI detection in ICUs.

Read paper
0 comments

No comments yet. Be the first.

Related papers

LatestFoundational
Cohort Study
odds ratio = 0.273
morphine was associated with a lower risk of delirium compared to fentanyl
Cohort Study
OR 2.22
being COVID-19 positive meant a higher chance of dying in the hospital
Cohort Study
OR: 2.172
sepsis-related clotting issues are associated with over twice the odds of hospital death
Cohort Study
0.84
the model performed well in identifying patients at higher risk for ventilator-associated pneumonia
Cohort Study
adjusted hazard ratio 1.917
Patients with leukoerythroblastosis had a 1.9-fold higher hazard of 90-day death.
Guideline
116mg/dL
lower alpha-1 antitrypsin protein levels than this trigger a two-step diagnosis
Review
80.6%
most reinforcement learning studies for sepsis treatment used the MIMIC database