Predicting Human Metabolic Rate Based on Multiple Physiological Parameters: A Pilot Study

Area of Science:

• Human physiology and thermal comfort
• Biomedical engineering and predictive modeling

Background:

• Accurate metabolic rate prediction is crucial for evaluating thermal comfort.
• Traditional methods are static and lack adaptability in dynamic environments.

Purpose of the Study:

• To investigate the correlation between physiological parameters (heart rate, mean arterial pressure, skin conductance level, ear temperature) and metabolic rate.
• To compare the performance of linear regression, support vector regression, and neural networks using various parameter combinations.
• To evaluate the effectiveness of a model driven solely by physiological parameters for metabolic rate prediction.

Main Methods:

• Recruited 36 healthy young participants for experiments involving nine exercise levels.
• Measured four physiological parameters and calculated metabolic rate using ISO 8996:2021 standards.
• Analyzed correlations, compared multi-parameter combinations, and evaluated model generalization using training and testing sets.

Main Results:

• Physiological parameters show sex-specific and interval-dependent relationships with metabolic rate, exhibiting nonlinear characteristics.
• Combining multiple physiological parameters significantly enhances prediction accuracy.
• A two-layer neural network model achieved superior predictive and generalization performance, though larger sample sizes are needed.

Conclusions:

• This preliminary study confirms the potential of using combined physiological parameters for metabolic rate prediction.
• Further research with larger sample sizes is recommended to refine models for practical application, especially under low metabolic rates.