Geometric distortion correction in gradient-echo imaging by use of dynamic time warping

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging
• Image Processing

Background:

• Local static field variations cause geometric distortion in Magnetic Resonance Imaging (MRI).
• The Chang and Fitzpatrick (CF) method corrects these distortions using two identical acquisitions with opposite read gradient polarities, previously applied to spin-echo imaging.

Purpose of the Study:

• To investigate the application of the CF method for correcting geometric distortion in gradient-echo imaging.
• To evaluate the effectiveness of a dynamic programming algorithm in conjunction with the CF method.
• To determine if the corrected images allow for local magnetic field extraction.

Main Methods:

• The Chang and Fitzpatrick (CF) method was adapted for gradient-echo imaging.
• A dynamic programming algorithm was employed to process the acquired data.
• Two gradient-echo acquisitions, differing only in read gradient polarity, were performed.

Main Results:

• The adapted CF method successfully corrected geometric distortion in gradient-echo imaging.
• The dynamic programming algorithm effectively mitigated edge artifacts common to the CF method.
• Local magnetic field information could be successfully extracted from the corrected images.

Conclusions:

• The Chang and Fitzpatrick method, when combined with a dynamic programming algorithm, is effective for correcting geometric distortion in gradient-echo MRI.
• This approach offers improved image quality by reducing artifacts.
• The method enables the retrieval of local magnetic field information, adding diagnostic value.