Related Experiment Video
Updated: May 11, 2025
09:36
Halogenated Agent Delivery in Porcine Model of Acute Respiratory Distress Syndrome via an Intensive Care Unit Type Device
Published on: September 24, 2020
2.7K
A machine learning-based severity stratification tool for high altitude pulmonary edema
Luobu Gesang1,2,3, Yangzong Suona4,5, Zhuoga Danzeng6,7
1High Altitude Medical Research Institute of Tibet Autonomous Region, 18 Linkuo North Road, Lhasa, 850000, China. KelsangNorbu@hotmail.com.
BMC Medical Informatics and Decision Making
|April 18, 2025
Summary
Identifying key predictors for High Altitude Pulmonary Edema (HAPE) severity is crucial. The random forest model accurately identifies severe HAPE cases, aiding prompt treatment and reducing mortality.
Area of Science:
- Altitude Medicine
- Pulmonary Physiology
- Machine Learning in Healthcare
Background:
- High Altitude Pulmonary Edema (HAPE) is a life-threatening condition.
- Prompt recognition of HAPE severity is critical for effective treatment and mortality reduction.
Purpose of the Study:
- To identify key predictors of High Altitude Pulmonary Edema (HAPE) severity.
- To develop and evaluate machine learning models for HAPE severity prediction.
- To assist clinicians in early identification of severely affected HAPE patients.
Main Methods:
- Analysis of 508 HAPE patients with 53 variables.
- Utilized Multinomail logistic regression, random forest, and decision tree models.
- Identified lung rales, sputum production, heart rate, and oxygen saturation as key predictors.
Main Results:
- The random forest model achieved the highest performance (77.94% accuracy, 0.86 AUC for severe cases).
- Lung rales, sputum production, heart rate, and oxygen saturation were significant predictors.
- The random forest model demonstrated high predictive accuracy across all HAPE severity levels.
Conclusions:
- A valuable screening tool for categorizing HAPE severity has been developed.
- The random forest model effectively aids in recognizing severe HAPE cases.
- This tool can facilitate prompt treatment and improve therapeutic approaches for HAPE.
