Evaluation of AI-Assisted Compressed Sensing for Accelerated Shoulder FACT MRI Sequences

Motivation: This study investigates the application of AI-assisted Compressed Sensing (ACS) acceleration in shoulder fat Analysis and Calculation Technique (FACT) sequences, assessing its impact on image quality across different tissues Goal(s): To evaluate the effect of ACS on image quality and quantitative accuracy when accelerating FACT sequences in shoulder. Approach: Twenty-one patients were scanned using a PET/MR system with both FACT and FACT ACS protocols. Quantitative parameters, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were compared across various anatomical structures. Results: The integration of FACT with ACS maintained image quality in most regions and significantly reduced scan times. Impact: The study demonstrates that ACS improves the clinical efficiency of FACT sequences in shoulder imaging by reducing scan times without affecting quantitative metrics or image quality.