Undersampled linogram trajectory for fast imaging (ULTI): experiments at 3 T and 7 T

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Image Reconstruction
• Medical Imaging Physics

Background:

• Undersampled data in MRI necessitates advanced reconstruction techniques.
• Parallel imaging methods accelerate data acquisition but can introduce artifacts.
• Evaluating novel sampling schemes is crucial for improving MRI performance.

Purpose of the Study:

• To investigate the performance of linogram acquisition for image reconstruction from undersampled MRI data.
• To compare linogram sampling with conventional radial sampling regarding image sharpness and artifacts.
• To assess the feasibility of linogram acquisition at high acceleration rates in both phantom and in vivo studies.

Main Methods:

• Analysis of the point spread function (PSF) for linogram sampling to assess image sharpness and artifacts.
• Implementation of generalized auto-calibrating partially parallel acquisition (PC-ASL) for linogram sampling.
• Comparison using simulations, phantom experiments at 3 Tesla, and in vivo human scanning at 7 Tesla.

Main Results:

• Linogram sampling demonstrated a sharper PSF and lower mean artifact power than radial sampling.
• Simulations showed lower reconstruction errors with linogram sampling.
• Phantom experiments preserved fine details and sharp edges with linogram sampling, unlike radial sampling which caused blurring.
• In vivo scans at 7T yielded high-quality anatomical images with linogram acquisition at high acceleration rates.

Conclusions:

• Linogram acquisition provides sharper images with fewer artifacts than radial sampling.
• It retains the benefits of radial sampling, including motion insensitivity and high acceleration rates.
• Linogram sampling is less susceptible to off-resonance artifacts, offering superior image quality in MRI.