Detection and Compensation of Periodic Motion in Magnetic Particle Imaging

Area of Science:

• Medical Imaging
• Biophysics
• Signal Processing

Background:

• Magnetic Particle Imaging (MPI) offers high temporal resolution but suffers from reduced spatial resolution due to low signal-to-noise ratios.
• Image reconstruction in MPI requires increased regularization, degrading spatial resolution, especially with low signal-to-noise ratio data.
• Block-wise signal averaging improves signal-to-noise ratio but compromises temporal resolution, hindering motion analysis.

Purpose of the Study:

• To develop a framework for averaging motion-corrupted MPI raw data to improve image quality.
• To address the challenge of synchronizing MPI acquisition frames with periodic physiological motion (respiration, heartbeat).
• To reduce motion artifacts in MPI images without requiring external navigator signals.

Main Methods:

• A novel framework is proposed for averaging motion-corrupted MPI raw data, considering periodic motion.
• The method involves shifting individual frames to align with the same state of motion and utilizes a binning technique for high-frequency motion.
• Motion frequency is determined directly from MPI raw data, eliminating the need for additional navigator signals.

Main Results:

• The proposed method successfully averages experimental data, demonstrating reduced motion artifacts on a motion phantom.
• Validation on in-vivo mouse data confirms the method's efficacy in compensating for heartbeat artifacts.
• The technique improves signal-to-noise ratio and spatial resolution in MPI images affected by physiological motion.

Conclusions:

• The developed framework effectively reduces motion artifacts in MPI by averaging corrupted raw data.
• This approach enhances image quality and allows for better analysis of dynamic processes in MPI.
• The method offers a practical solution for improving MPI performance in the presence of physiological motion.