Near-metal MRI Using PETRA With Extended Phase Encoding and Compressed Sensing

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging (MRI)
• Biomedical Engineering

Background:

• Metallic implants in patients pose significant challenges for MRI due to artifacts.
• Existing MRI sequences often struggle to provide clear images near implants, limiting diagnostic capabilities.
• Compressed Sensing (CS) techniques offer potential solutions for faster and artifact-reduced MRI.

Purpose of the Study:

• To adapt and evaluate the Compressed Sensing Parallel Imaging with Echo Train Acquisition (csPETRA) sequence for MRI near metallic orthopedic implants.
• To investigate the impact of an extended phase encoding in csPETRA on image quality and acquisition time.
• To compare the performance of csPETRA with conventional sequences like TSE with VAT and SEMAC.

Main Methods:

• Simulations using a 3D orthopedic implant model to test csPETRA with regular and extended phase encoding.
• Acquisition of k-space samples from 3.2% to 6.5% within a 7±1 minute scan time.
• Comparison with conventional MRI sequences (TSE with VAT and SEMAC) in phantom studies and in vivo patient scans.

Main Results:

• csPETRA demonstrated a 40% reduction in geometric distortions compared to SEMAC-VAT.
• An extended phase encoding (6.5% SPI coverage) improved artifact size by 20% over a lower coverage (3.2%).
• A 16-fold acceleration in csPETRA reduced scan time to 8 minutes, with a 2.5-fold reduction in SAR compared to SE techniques.

Conclusions:

• csPETRA with optimized phase encoding effectively reduces MRI artifacts near metallic implants without prolonging scan times.
• The technique shows promise for niche applications like high-resolution field mapping, implant assessment, and imaging patients with SAR constraints.
• Further validation is needed to establish csPETRA's clinical utility, particularly regarding soft-tissue contrast mechanisms.