Enhance-then-Synthesize: A Deep Learning Acceleration Pipeline for Spinal MRI, Reducing Scan Time by 60%

Motivation: The MRI spine protocol consists of multiple sequences for routine assessment but takes over 5 minutes, often resulting in inefficient, motion-affected image quality. Goal(s): We proposed a deep learning(DL)-based pipeline that boosts speed by 60% and evaluated its impact on image quality and diagnostic performance. Approach: Using DL-based image enhancement and generative algorithms, we generated four high-quality sequences from three low-quality fast scan sequences. These sequences were then compared to the SOC for image quality, similarity, and diagnostic consistency. Results: The image quality of the DL-generated sequences surpasses that of the SOC, demonstrates diagnostic interchangeability, and reduces scan time by 60%. Impact: The DL-based enhance-then-synthesize pipeline reduced the scanning time for spine MRI protocols up to 60%, while delivering image quality comparable to or higher than that of the standard scanning sequence. The generated sequences proved to be viable alternatives for diagnosis.