Key result
Bagged trees machine learning predicts mild acute mountain sickness with ~100% accuracy.
Why the study?
Recent studies on acute mountain sickness have used fixed-location and fixed-time measurements, lacking real-time assessment of environmental and physiological variables to forecast AMS development.
Does machine learning multivariate analysis of physiological and environmental parameters improve the prediction of acute mountain sickness in hikers compared to single variable analysis?
Comparison
Machine learning multivariate analysis vs single variable analyses
Design
Observational study with regression and classification machine learning analyses
Authors
Loading...
Real-time ML model aids AMS risk prediction; leaves open prospective validation before clinical adoption.
Observational (n=32)
Does machine learning multivariate analysis of physiological and environmental parameters improve the prediction of acute mountain sickness in hikers compared to single variable analysis?
Effect estimate: AUC 1
A machine learning model incorporating real-time physiological and environmental variables accurately predicts acute mountain sickness risk in hikers.
Wei et al. (2021) conducted an observational in Acute mountain sickness (n=32). Physiological and environmental parameters (altitude, temperature, pressure, humidity, climbing speed, heart rate, SpO2, HRV) was evaluated on Prediction of mild acute mountain sickness (Lake Louise Score ≥ 3) using a bagged trees classifier (AUC 1). A bagged trees machine learning classifier incorporating multiple physiological and environmental variables predicted the onset of mild acute mountain sickness with an accuracy of 0.998 and an AUC of 1.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: