Best fast MRI protocols for the knee: advantages and limitations.

Area of Science:

• Radiology
• Medical Imaging
• Magnetic Resonance Imaging

Background:

• Knee MRI is crucial for musculoskeletal diagnostics but historically involves long scan times.
• Technological advancements are revolutionizing MRI acquisition speed.
• Deep learning (DL) reconstruction offers new possibilities for accelerating MRI.

Purpose of the Study:

• To review modern acceleration strategies in knee MRI.
• To summarize evidence on diagnostic performance and image quality of accelerated techniques.
• To discuss practical implementation and future directions for fast knee MRI.

Main Methods:

• Utilizing parallel imaging (PI), simultaneous multi-slice (SMS) acquisition, and compressed sensing (CS).
• Employing deep learning (DL)-based reconstruction for higher acceleration factors.
• Investigating accelerated 3D TSE techniques like CAIPIRINHA-accelerated SPACE sequences.

Main Results:

• Sixfold PIxSMS-accelerated DL protocols achieve excellent diagnostic performance and image quality in under five minutes.
• Accelerated 3D TSE techniques enable practical, high-resolution isotropic 3D imaging.
• DL reconstruction shows promise for acquiring multi-contrast isotropic 3D datasets rapidly.

Conclusions:

• Modern acceleration strategies, especially with DL, drastically reduce knee MRI scan times.
• Fast knee MRI protocols are becoming clinically practical, enhancing workflow efficiency.
• Future developments in DL and workflow optimization will further advance rapid, comprehensive knee MRI.