Temporal AI-assisted compressed sensing for high-resolution, motion-robust small-bowel MR enterography without...

Area of Science:

• Medical Imaging
• Artificial Intelligence in Radiology
• Gastrointestinal Imaging

Background:

• Magnetic Resonance Enterography (MRE) is crucial for evaluating small bowel diseases.
• Routine MRE often requires antiperistaltic agents to minimize motion artifacts.
• Improving MRE acquisition efficiency and image quality without medication is a clinical need.

Purpose of the Study:

• To evaluate the efficacy of temporal AI-assisted compressed sensing (tACS) for MRE.
• To determine if tACS can achieve high-resolution, motion-robust MRE without antiperistaltic agents.
• To assess improvements in acquisition efficiency and motility visualization with tACS.

Main Methods:

• Prospective single-center study comparing routine MRE with tACS MRE on a 3.0-T scanner.
• No antiperistaltic agents were used in either protocol.
• Key outcomes included imaging efficiency (frames/min), image quality scores, sharpness, signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and ileocecal valve (ICV) motility visualization.

Main Results:

• tACS improved imaging efficiency four-fold (105.4 vs 26.6 frames/min; p < 0.001).
• Subjective image quality was significantly better with tACS (p < 0.01), with improved sharpness, SNR, and CNR.
• tACS provided superior visualization of ICV motility compared to routine Cine sequences (p < 0.001).

Conclusions:

• tACS enables high-resolution, motion-robust MRE without antiperistaltic agents.
• The technique significantly enhances image quality, acquisition efficiency, and motility visualization.
• tACS has the potential to simplify MRE workflows and improve patient comfort.