Accelerated T2-Weighted TSE Imaging of the Prostate Using Deep Learning Image Reconstruction: A Prospective...

Area of Science:

• Radiology
• Medical Imaging
• Artificial Intelligence

Background:

• Multiparametric MRI (mpMRI) is standard for prostate cancer evaluation.
• Deep learning image reconstruction (DLR) shows promise for accelerating MRI scans.

Purpose of the Study:

• To assess the impact of DLR on T2-weighted TSE prostate MRI.
• To evaluate image quality, diagnostic confidence, PI-RADS, and T2 scoring with DLR compared to standard imaging.

Main Methods:

• Prospective study of 60 patients undergoing 3T mpMRI for prostate cancer.
• Acquisition of standard T2 TSE (T2S) and DLR-enhanced T2 TSE (T2DLR) sequences.
• Independent radiologist evaluation of image quality, artifacts, conspicuity, and scoring using a Likert scale.

Main Results:

• DLR reduced acquisition time by over 60% (10:21 min vs. 3:50 min).
• DLR significantly improved noise levels, reduced artifacts, and enhanced overall image quality and lesion conspicuity.
• Both readers found superior image quality and lesion conspicuity with T2DLR compared to T2S.

Conclusions:

• T2-weighted TSE prostate MRI with DLR is feasible, reducing scan time significantly.
• DLR offers substantial improvements in image quality and diagnostic confidence for prostate cancer evaluation.