Quantitative multislice and jointly optimized rapid CEST for in vivo whole-brain imaging

Area of Science:

• Magnetic Resonance Imaging
• Quantitative Imaging
• Biomedical Engineering

Background:

• Chemical Exchange Saturation Transfer (CEST) MRI is a powerful technique for assessing tissue properties.
• Multislice CEST sequences often suffer from reduced sensitivity and efficiency.
• Optimizing acquisition schedules is crucial for improving quantitative CEST imaging.

Purpose of the Study:

• To develop a novel multislice CEST pulse sequence and schedule optimization framework.
• To enhance sensitivity and reduce scan time for quantitative CEST MRI.
• To enable accurate and reproducible whole-brain quantitative CEST imaging.

Main Methods:

• A deep learning framework was employed for simultaneous optimization of scan parameters and slice order.
• The optimized sequence was evaluated through numerical simulations and in vivo scans of healthy subjects.
• Quantitative metrics including scan efficiency, error, and reproducibility (Lin's CCC) were assessed.

Main Results:

• The optimized multislice sequence demonstrated superior scan efficiency and reduced error compared to random and single-slice schedules.
• In vivo quantitative CEST values agreed well with the single-slice sequence and previous studies.
• High test-retest and intersubject reproducibility were achieved, with average WM/GM CCC values of 0.8151/0.7779 and 0.7792/0.7191, respectively.

Conclusions:

• A robust framework for multislice CEST schedule optimization and pulse sequence design was successfully developed.
• The proposed approach facilitates accurate, reproducible, and efficient whole-brain quantitative CEST imaging.
• This advancement holds promise for clinical applications requiring rapid and reliable tissue characterization.