Improving apparent diffusion coefficient accuracy on a compact 3T MRI scanner using gradient nonlinearity correction.

Area of Science:

• Magnetic Resonance Imaging
• Diffusion-Weighted Imaging

Background:

• Gradient nonlinearity (GNL) introduces bias in apparent diffusion coefficients (ADCs) during diffusion-weighted imaging.
• A gradient nonlinearity correction (GNLC) method, previously validated for whole-body systems, was assessed for its efficacy on a novel compact 3T (C3T) scanner.
• The C3T scanner's asymmetrical design results in more complex GNL, necessitating specific evaluation.

Purpose of the Study:

• To evaluate the effectiveness of GNLC in improving ADC quantification accuracy on the C3T scanner.
• To compare ADC values obtained with and without GNLC against established literature values.

Main Methods:

• Utilized a diffusion quality control phantom with varying polyvinylpyrrolidone concentrations.
• Acquired imaging data from the C3T scanner and two commercial 3T scanners.
• Analyzed retrospective patient data (n=12) including cerebrospinal fluid and white matter regions.

Main Results:

• Corrected ADC values for the phantom were within 2.5% of reference values across all concentrations.
• Post-GNLC, ADC values for cerebrospinal fluid and white matter in patients closely matched literature values (within 1% and 6%, respectively).
• Significant improvement (P < 0.05) in ADC accuracy was observed after applying GNLC compared to uncorrected data.

Conclusions:

• The C3T scanner exhibits more pronounced GNL effects due to its asymmetric gradient configuration.
• The implemented GNLC effectively compensates for these GNL-induced biases.
• GNLC ensures accurate ADC quantification on the C3T scanner, aligning results with established literature data.