Related Experiment Video
Updated: Sep 21, 2025
04:04
Asthma Detection Research Based on Voice Signal Processing and Machine Learning
Published on: July 22, 2025
437
Using machine learning to determine the correlation between physiological and environmental parameters and the
Chih-Yuan Wei1, Ping-Nan Chen2,3, Shih-Sung Lin4
1Graduate Institute of Life Sciences, National Defense Medical Center, No.161, Sec. 6, Minquan E. Rd., Neihu Dist., Taipei, 11490, Taiwan.
BMC Bioinformatics
|June 1, 2022
Summary
This study developed a machine learning model to predict acute mountain sickness (AMS) risk in real-time using physiological and environmental data. The model achieved high accuracy, enabling timely warnings for hikers to prevent AMS onset.
Area of Science:
- Altitude sickness research
- Machine learning applications in healthcare
- Environmental physiology
Background:
- Traditional acute mountain sickness (AMS) studies rely on fixed measurements.
- Real-time data collection is crucial for understanding AMS development.
- Developing predictive models can aid in AMS prevention.
Purpose of the Study:
- To measure real-time environmental and physiological variables during high-altitude ascent.
- To develop a machine learning-based AMS risk evaluation model.
- To forecast AMS onset for preventative measures.
Main Methods:
- Recruited 32 participants (25 men, 7 women) hiking from 2300m to 3275m.
- Collected real-time physiological (heart rate, SpO2, HRV) and environmental data.
- Applied regression and classification machine learning analyses to predict AMS risk using Lake Louise Scores (LLS).
Main Results:
- A combined model incorporating all variables achieved an R² of 0.62.
- The bagged trees classifier demonstrated high performance: 0.999 sensitivity, 0.994 specificity, 0.998 accuracy, and a 1 AUC.
- Multivariate analysis using machine learning significantly improved AMS prediction accuracy.
Conclusions:
- Machine learning multivariate analysis offers superior AMS prediction accuracy compared to single-variable analyses.
- The developed AMS evaluation model can inform the creation of wearable devices for real-time risk warnings.
- This approach facilitates proactive management and prevention of acute mountain sickness.
