Cartesian MR fingerprinting in the eye at 7T using compressed sensing and matrix completion-based reconstructions

Area of Science:

• Medical Imaging
• Biophysics

Background:

• Accurate quantification of tissue relaxation times is crucial for diagnosing eye conditions.
• Current MRI techniques can be time-consuming and susceptible to motion artifacts, particularly in the eye.

Purpose of the Study:

• To assess the feasibility of Magnetic Resonance Fingerprinting (MRF) for rapid relaxation time quantification in the human eye at 7 Tesla.
• To establish a framework for MRF data acquisition and processing for future characterization of eye tumors.

Main Methods:

• A single-element receive coil MRF approach with Cartesian sampling was employed.
• Undersampling combined with compressed sensing (CS) and matrix completion (MC) reconstruction techniques were investigated.
• MRF parameter maps were evaluated through simulations and experiments in healthy volunteers and one uveal melanoma patient.

Main Results:

• Matrix completion (MC) reconstruction demonstrated superior accuracy and allowed for higher undersampling factors compared to compressed sensing (CS).
• Scan time was significantly reduced from over 7 minutes to approximately 1 minute 16 seconds with MC, without compromising image quality.
• Measured relaxation times in healthy volunteers aligned with literature values, and MRF successfully differentiated tumor from healthy tissue in a patient.

Conclusions:

• Cartesian-based MRF is a feasible technique for rapid relaxation time quantification in the human eye at 7T.
• Matrix completion enables substantial scan time reduction and mitigation of motion artifacts, enhancing patient comfort and diagnostic potential for eye tumor characterization.