ε-Tube regression: a new method for motion artifact reduction

Area of Science:

• Biomedical Engineering
• Signal Processing
• Physiological Monitoring

Background:

• Motion artifacts are a common challenge in physiological signal acquisition.
• Accurate physiological signal analysis requires effective artifact removal techniques.

Purpose of the Study:

• To introduce a new regression method, ε-tube regression (ε-TR), for motion artifact reduction.
• To demonstrate the efficacy of ε-TR in removing motion artifacts from physiological signals, specifically arm impedance signals.

Main Methods:

• Developed ε-tube regression (ε-TR), a novel approach that forms a 'tube' around the data.
• The method models only the motion artifact, distinguishing it from the target physiological signal.
• Incorporated information about the motion artifact's shape into the approximation process.

Main Results:

• ε-tube regression (ε-TR) successfully modeled and isolated motion artifacts.
• The method effectively removed motion artifacts from impedance signals measured on the arms.
• The target physiological signal remained uncompromised after artifact removal.

Conclusions:

• ε-tube regression (ε-TR) is a promising technique for motion artifact reduction in physiological signals.
• This method offers a robust solution for improving the quality of impedance measurements.
• Further applications of ε-TR in other physiological signals are warranted.