A Respiratory Motion Estimation Method Based on Inertial Measurement Units for Gated Positron Emission Tomography

Area of Science:

• Biomedical Engineering
• Medical Imaging Physics
• Wearable Sensor Technology

Background:

• Respiratory motion significantly impacts medical imaging quality, particularly in Positron Emission Tomography (PET).
• Accurate respiratory motion estimation is crucial for improving image resolution and quantitative accuracy in PET scans.
• Current methods for respiratory motion estimation can be complex or lack quantitative precision.

Purpose of the Study:

• To introduce a novel method for estimating respiratory motion using commercial microelectromechanical systems (MEMS) inertial measurement units (IMUs).
• To apply this method for amplitude gating in PET imaging and compare its performance against a standard clinical technique.
• To enable quantitative estimation of both magnitude and phase of respiratory motion from chest and abdomen.

Main Methods:

• Utilized MEMS-based IMUs to capture chest and abdominal movements.
• Developed algorithms to detect respiratory cycles and estimate their length, magnitude, and phase.
• Compared the novel IMU-based method against a clinically used respiratory motion estimation technique in a cohort of patients.

Main Results:

• The IMU-based method demonstrated state-of-the-art accuracy in detecting respiratory cycles and estimating their lengths.
• Achieved an absolute breathing rate error of 0.44 ± 0.23 1/min and an absolute amplitude error of 0.24 ± 0.09 cm in subjects with acute myocardial infarction.
• This is the first IMU-based method using commercial MEMS devices capable of quantitative magnitude and phase estimation.

Conclusions:

• The novel MEMS IMU method offers a simplified and accurate approach to respiratory motion estimation for PET imaging.
• The technique has the potential to enhance PET imaging logistics and enable advanced organ motion studies.
• This method provides a robust platform for quantitative respiratory motion analysis using accessible wearable sensors.