Quantitative molecular imaging using deep magnetic resonance fingerprinting

Area of Science:

• Biomedical Imaging
• Artificial Intelligence in Medicine
• Quantitative MRI

Background:

• Saturation transfer-based molecular MRI is complex and slow for clinical use.
• Existing methods struggle with quantitative molecular biomarker extraction.
• Deep learning offers a potential solution to these challenges.

Purpose of the Study:

• To define a complete protocol for quantitative molecular MRI using deep MRF.
• To address limitations of traditional saturation transfer MRI.
• To provide a rapid and quantitative framework for molecular imaging.

Main Methods:

• Developed a deep learning-based saturation transfer MRF protocol.
• Included sample preparation, acquisition design, model training, and deep reconstruction.
• Optimized chemical exchange saturation transfer (CEST) and magnetization transfer (MT) imaging.

Main Results:

• Demonstrated deep MRF's performance in cancer monitoring, brain myelin imaging, and pH quantification.
• Provided a framework suitable for graduate-level users.
• The protocol can be adapted for various pathologies like neurodegeneration and cardiac disease.

Conclusions:

• Deep MRF provides a quantitative and rapid solution for molecular MRI.
• This protocol facilitates in vivo imaging of molecular biomarkers.
• The open-source code and data enable further research and application development.