Energy Expenditure Estimation From Respiratory Magnetometer Plethysmography: A Comparison Study

Area of Science:

• Biomedical Engineering
• Exercise Physiology
• Wearable Technology

Background:

• Accurate physical activity (PA) and energy expenditure (EE) quantification is crucial for health monitoring.
• Traditional reference methods for EE estimation are often expensive and impractical for widespread use.
• Development of lightweight, cost-effective portable devices is needed to overcome these limitations.

Purpose of the Study:

• To comparatively evaluate the performance of portable devices, including Respiratory Magnetometer Plethysmography (RMP), for estimating EE across a range of physical activity intensities.
• To assess the efficacy of machine learning algorithms (Artificial Neural Network and Support Vector Regression) for EE estimation using sensor data.
• To compare different validation approaches for the developed models.

Main Methods:

• Fifteen healthy subjects performed nine different sedentary and physical activities.
• Subjects were monitored using an accelerometer, heart rate (HR) monitor, RMP device, and a gas exchange system.
• Artificial Neural Network (ANN) and Support Vector Regression (SVR) models were developed using sensor features, with comparisons of leave-one-out, 10-fold cross-validation, and subject-specific validation.

Main Results:

• The RMP device demonstrated superior EE estimation accuracy compared to accelerometer or HR monitor used individually.
• Combining RMP and HR data significantly enhanced the performance of EE estimation.
• The RMP device proved reliable for estimating EE across diverse physical activity intensities, from sedentary to high-intensity exercise.

Conclusions:

• Respiratory Magnetometer Plethysmography (RMP) is a promising portable technology for accurate energy expenditure estimation.
• Integration of RMP with heart rate monitoring offers improved accuracy for physical activity energy expenditure assessment.
• The RMP device provides a reliable and cost-effective solution for quantifying energy expenditure in various real-world scenarios.