Free-Breathing Hepatic Oxygen Extraction Fraction (OEF) Mapping Using Radial GESSE

Area of Science:

• Medical Imaging
• Biophysics
• Hepatology

Background:

• Noninvasive hepatic oxygenation assessment is crucial for liver disease evaluation.
• Current methods are limited by single-slice imaging and breath-holding requirements.
• Developing motion-robust, whole-liver techniques is essential for accurate diagnosis.

Purpose of the Study:

• To develop and evaluate a motion-robust technique for whole-liver oxygen extraction fraction (OEF) mapping.
• To utilize a radial gradient-echo sampling of spin-echo (rGESSE) sequence combined with an artificial neural network (ANN).
• To enable noninvasive, volumetric hepatic oxygenation assessment under free-breathing conditions.

Main Methods:

• A multi-slice rGESSE sequence with radial sampling was used on a 1.5T MRI system.
• Seven healthy volunteers underwent scanning under free-breathing conditions.
• An artificial neural network (ANN) was employed for quantifying OEF, deoxygenated blood volume (DBV), and R2 from simulated qBOLD signal models.

Main Results:

• Whole-liver OEF maps were successfully generated in all volunteers during free breathing.
• The mean hepatic OEF was 55.75% ± 6.88%, and mean DBV was 0.568 ± 0.039.
• ANN-based fitting demonstrated superior stability and artifact suppression compared to standard regression.

Conclusions:

• The rGESSE sequence combined with ANN analysis enables feasible free-breathing, whole-liver OEF mapping.
• This approach offers noninvasive, volumetric hepatic oxygenation assessment with enhanced motion robustness.
• The technique holds potential for clinical applications in diagnosing and monitoring liver diseases.