Sci-Fri AM: YIS-01: Comprehensive MR distortion correction: Phantom validation and in-vivo application

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Image Processing

Background:

• Magnetic Resonance (MR) imaging offers superior soft-tissue contrast for diagnostics.
• Geometric uncertainties in MR images limit their utility in treatment planning.
• Growing interest in MR-based treatment planning necessitates accurate geometric correction.

Purpose of the Study:

• To systematically quantify machine (gradient non-linearity, B0 inhomogeneity) and patient (susceptibility, chemical shift) distortions in MR images.
• To develop and validate a distortion correction method for enhancing MR image accuracy in treatment planning.

Main Methods:

• Gradient non-linearities measured using a 3D grid phantom.
• B0 inhomogeneity and susceptibility distortions measured using double gradient echo scans.
• Chemical shift distortion evaluated using fat and water inserts; intensity correction validated using a uniformity phantom.

Main Results:

• Maximum absolute distortion reduced from 5.4 mm to 1.0 mm; mean distortion reduced from 1.7 ± 1.4 mm to 0.4 ± 0.2 mm for B0 inhomogeneity and susceptibility.
• Significant improvement in fat signal displacement after chemical shift distortion correction.
• Intensity correction demonstrated necessity through comparison of corrected and uncorrected image profiles.

Conclusions:

• The developed systematic approach effectively quantifies and corrects multiple MR image distortions.
• Validated correction techniques improve geometric accuracy, crucial for MR-guided treatment planning.
• The method was successfully implemented on volunteer brain images (GE and multi-shot EPI).