Accelerated CMR using zonal, parallel and prior knowledge driven imaging methods.

Area of Science:

• Cardiovascular Magnetic Resonance (CMR) Imaging
• Medical Imaging Acceleration Techniques

Background:

• Cardiovascular magnetic resonance (CMR) applications face challenges balancing spatiotemporal resolution with scan time.
• Data undersampling is a key strategy for enhancing scan efficiency in CMR.

Purpose of the Study:

• To review and discuss three primary approaches for accelerating CMR imaging.
• To highlight the benefits and limitations of various acceleration techniques.

Main Methods:

• Discussion of zonal imaging, particularly its applications in coronary imaging and flow measurements.
• Detailed review of parallel imaging (PI) and its widespread use in functional, perfusion, viability, and coronary imaging.
• Exploration of prior knowledge-driven methods, often combined with PI, for advanced acceleration.

Main Results:

• Parallel imaging routinely achieves reduction factors of 2-3, with potential for >3-4 with large coil arrays.
• Prior knowledge-driven methods have demonstrated 5-8 fold accelerations in cine and dynamic CMR, with favorable image quality compared to PI alone.
• Fundamental limits of coil encoding exist, necessitating careful consideration of acceleration trade-offs and artifact potential.

Conclusions:

• Accelerated CMR imaging is vital for clinical practice, with PI and prior knowledge methods offering significant speed-ups.
• Understanding and mitigating artifacts, such as spatially varying noise in PI and temporal fidelity issues, is essential for reliable results.
• Future developments in coil technology and imaging protocols will continue to advance CMR acceleration.