4D MR imaging using robust internal respiratory signal.

Area of Science:

• Medical Imaging
• Biophysics

Background:

• Four-dimensional (4D) magnetic resonance (MR) imaging requires accurate sorting of respiratory motion.
• Traditional methods like respiratory bellows can be prone to errors.

Purpose of the Study:

• To investigate the feasibility of using internal respiratory (IR) surrogates for sorting 4D MR images.
• To develop a novel method for deriving IR signals from MR data.

Main Methods:

• Acquisition of fast 2D cine MR images to construct 4D volumes.
• Retrospective sorting using derived IR signals.
• Utilizing low-frequency Fourier components and anterior body boundary as potential IR surrogates.
• Employing a clustering algorithm to identify optimal IR surrogates and derive the IR signal.
• Comparison of derived IR signals with respiratory bellows in healthy volunteers.

Main Results:

• 99% agreement between the proposed IR signal and respiratory bellows signals.
• Average end-inspiration time difference of 0.18s for matching signals.
• Identified instances where the IR signal potentially corrected faulty bellows signals.
• Sorted 4D MR images exhibited minimal artifacts and demonstrated clinical potential.

Conclusions:

• The proposed method using internal respiratory surrogates is a feasible and effective alternative for sorting 4D MR images.
• This technique offers improved accuracy and potential for correcting erroneous external respiratory monitoring.
• The sorted 4D MR images show promise for clinical applications.