Deep learning reconstruction based rapid T2-weighted imaging improved the image quality for primary rectal cancer staging

Motivation: High-resolution T2-weighted imaging (T2WI) is routinely used for non-invasively staging primary rectal cancer (RC). However, the RC T2WI was lengthy in clinic using standard imaging technique. Goal(s): Exploring the feasibility of reducing RC T2WI scan time by applying deep learning reconstruction (DLR). Approach: 52 primary rectal cancer patients were enrolled.Each was imaged with standard clinical T2WI (cT2WI) and DLR based rapid T2WI (rT2WI). Results: Compared to cT2WI, DLR based rT2WI reduced 2/3 imaging time, presented with better image quality to more accurately delineate tumor anatomy, and relatively improved the diagnostic performance of primary RC TN staging. Impact: The application of DLR would be beneficial for rectum T2WI in terms of shorten scan time, improved tumor boundary delineation, and better node characteristic presentation which are important for primary rectal cancer TN staging.