Accelerating CEST MRI through complementary undersampling and multi-offset transformer reconstruction

Area of Science:

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

Background:

• Chemical Exchange Saturation Transfer (CEST) MRI offers valuable in vivo molecular information.
• Conventional CEST MRI requires lengthy acquisition times due to multiple frequency offset scans.

Purpose of the Study:

• To develop and validate a hybrid k-space and image domain strategy for accelerating CEST MRI.
• To enhance the efficiency and applicability of CEST MRI for broader clinical and pre-clinical use.

Main Methods:

• Implemented a complementary k-space undersampling strategy across adjacent frequency offsets.
• Developed a multi-offset transformer reconstruction network incorporating data consistency and coil combination layers.
• Validated the method on rodent and human brain CEST MRI data acquired on 3T scanners.

Main Results:

• The hybrid acceleration strategy significantly improved reconstruction accuracy and image fidelity compared to state-of-the-art methods.
• CEST maps, including amide proton transfer (APT) and relayed nuclear Overhauser enhancement (rNOE), demonstrated close agreement with fully-sampled data.
• The method proved effective on both pre-clinical and clinical multi-coil CEST MRI datasets.

Conclusions:

• The proposed hybrid acceleration technique effectively reduces CEST MRI scan times without compromising image quality or diagnostic accuracy.
• This approach facilitates wider adoption of advanced CEST MRI techniques for in vivo molecular imaging.