A framework for rapid, repeatable, and high-fidelity whole-brain multi-pool CEST imaging at 3 T

Area of Science:

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

Background:

• Chemical Exchange Saturation Transfer (CEST) imaging offers metabolic insights but faces challenges like long scan times and confounding factors.
• Developing rapid and accurate CEST techniques is crucial for clinical translation and broader application in neuroscience.

Purpose of the Study:

• To create and validate a framework for fast, accurate, and repeatable whole-brain, multi-pool CEST imaging at 3 Tesla.
• To address limitations of long acquisition times and confounding factors in quantitative CEST.

Main Methods:

• Optimized a single-shot 3D True FISP sequence for whole-brain multi-pool CEST acquisition.
• Implemented rapid B0, B1, and T1 mapping with a dual-echo modified four-angle method.
• Developed a neural network for B1 correction and utilized the AREX metric for T1 and MT correction.

Main Results:

• Achieved high-quality, whole-brain CEST images in ~9 minutes with minimal artifacts.
• Demonstrated significant reduction in coefficient of variation (CV) for MTR_LD and APT_AREX after B1 and AREX correction.
• Showcased excellent agreement between neural network B1 correction and conventional methods (ICC > 0.95) and high test-retest repeatability (average CV < 10% for 95/96 regions).

Conclusions:

• Successfully developed and validated a rapid, robust framework for quantitative whole-brain multi-pool CEST imaging.
• The integrated approach overcomes key barriers to clinical translation for CEST imaging.
• Enables reliable metabolic imaging for diverse brain pathologies.