Assessing Respiratory Activity by Using IMUs: Modeling and Validation

Area of Science:

• Biomedical Engineering
• Respiratory Physiology

Background:

• Accurate estimation of respiratory parameters is crucial for monitoring patient health.
• Inertial measurement units (IMUs) offer a potential non-invasive method for respiratory monitoring.

Purpose of the Study:

• To explore novel IMU-based strategies for estimating respiratory parameters in healthy adults.
• To identify a minimal set of IMUs and signals for accurate respiratory monitoring.

Main Methods:

• Simultaneous collection of chest wall kinematics using 9 IMUs and 45 reflective markers.
• Analysis of inertial kinematics to develop models matching tidal volume measured by optoelectronic plethysmography.
• Leave-one-out cross-validation for model tuning and accuracy assessment.

Main Results:

• Two models were identified: Model 1 (3 IMUs) and Model 2 (2 IMUs).
• Both models accurately estimate respiratory rate (error < 1.5%).
• Model 2 provides better accuracy for respiratory phase duration (error < 5%), while Model 1 excels in tidal volume estimation (error < 5%).

Conclusions:

• IMU-based strategies can accurately estimate key respiratory parameters.
• Minimal IMU configurations can achieve reliable respiratory monitoring.
• Further research can enhance the accuracy and applicability of these IMU-based methods.