Tri-axial high-resolution acceleration for oxygen uptake estimation: Validation of a multi-sensor device and a novel...

Area of Science:

• Exercise Physiology
• Biomedical Engineering
• Wearable Technology

Background:

• Accurate measurement of oxygen uptake (V̇O2) is crucial for exercise physiology and clinical assessments.
• Indirect calorimetry is the gold standard but is resource-intensive and limits real-world application.
• Wearable sensors offer a promising alternative for continuous V̇O2 monitoring.

Purpose of the Study:

• To validate a multi-sensor chest-strap system against indirect calorimetry.
• To introduce and evaluate the total-acceleration-variability (TAV) method for analyzing accelerometer data.
• To develop and assess linear regression models for predicting V̇O2 from TAV-processed multi-sensor data.

Main Methods:

• Validation of a multi-sensor chest-strap device.
• Application of the total-acceleration-variability (TAV) method to high-resolution accelerometer data.
• Development of linear regression models to predict oxygen uptake (V̇O2).

Main Results:

• Strong individual correlations between observed and TAV-predicted V̇O2 (mean r = 0.94).
• Low bias in TAV-predicted V̇O2 compared to indirect calorimetry (1.5 mL·min(-1)·kg(-1)).
• Confidence intervals indicate potential variability when using single-model predictions.

Conclusions:

• The multi-sensor chest-strap with TAV analysis provides a valid method for predicting V̇O2.
• While accurate on average, caution is advised when using single-model predictions as a direct surrogate for V̇O2.
• This technology holds potential for non-laboratory based V̇O2 assessment.