Single point imaging with radial acquisition and compressed sensing.

Area of Science:

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

Background:

• The Pointwise Encoding Time Reduction with Radial Acquisition (PETRA) sequence is crucial for high-resolution MRI of tissues with ultra-short T2 values.
• Accelerating MRI acquisition times is essential for clinical feasibility and patient comfort.
• Compressed sensing (CS) is a promising technique for accelerating MRI scans by reconstructing images from undersampled data.

Purpose of the Study:

• To accelerate the PETRA sequence using compressed sensing (csPETRA) for ultra-short T2 imaging.
• To evaluate the impact of acceleration on image quality and diagnostic accuracy in knee MRI.
• To reduce scan times while preserving anatomical detail and minimizing artifacts.

Main Methods:

• Compressed sensing was integrated into the PETRA sequence (csPETRA) to enable random undersampling of k-space data.
• Acceleration factors (Acc) ranging from 4 to 32 were applied.
• Phantom and in vivo knee images from six volunteers were acquired at 3T and compared to fully sampled (Acc=1) data for structural similarity and normalized-mean-square-error.

Main Results:

• No aliasing artifacts were observed even at Acc=16, with preserved anatomical details compared to fully sampled data.
• Normalized-mean-square-error remained below 1% at Acc=16.
• Single point imaging acquisition time was reduced from 165s to 10s, decreasing total scan time from 7.8 to 5.2 minutes at Acc=16.

Conclusions:

• csPETRA enables ultra-short T2 imaging of the knee joint within clinically acceptable scan times.
• The technique maintains the image quality of the original non-accelerated PETRA sequence.
• Accelerated csPETRA at Acc=16 provides comparable diagnostic quality to fully sampled data for knee imaging.