Your new experience awaits. Try the new design now and help us make it even better

ORIGINAL RESEARCH article

Front. Radiol.

Sec. Artificial Intelligence in Radiology

Volume 5 - 2025 | doi: 10.3389/fradi.2025.1695043

High-resolution deep learning reconstructed T2-weighted imaging for improvement of image quality and extraprostatic extension assessment in prostate MRI

Provisionally accepted
Sebastian  GassenmaierSebastian Gassenmaier1*Franziska  Katharina StaberFranziska Katharina Staber1Stephan  UrsprungStephan Ursprung1Judith  HerrmannJudith Herrmann1Sebastian  WernerSebastian Werner1Andreas  LinggAndreas Lingg1Lisa  C AdamsLisa C Adams2Haidara  AlmansourHaidara Almansour1Konstantin  NikolaouKonstantin Nikolaou1Saif  AfatSaif Afat1
  • 1Eberhard Karls Universitat Tubingen, Tübingen, Germany
  • 2Klinikum rechts der Isar der Technischen Universitat Munchen, Munich, Germany

The final, formatted version of the article will be published soon.

Purpose: To evaluate the impact of high-resolution T2-weighted imaging (T2HR) including deep learning image reconstruction (DLR), on image quality, lesion delineation, and extraprostatic extension (EPE) assessment in prostate multiparametric MRI (mpMRI). Materials and Methods: This retrospective study included 69 patients who underwent mpMRI of the prostate on a 3T scanner with DLR between April 2023 and March 2024. Routine mpMRI protocols adhering to PI-RADS v2.1 were used, including additionally a T2HR sequence (2 mm slice thickness, 4:31 min versus 4:12 min for standard T2 (T2S)). Image datasets were evaluated by two radiologists using a Likert scale ranging from 1 – 5 with 5 being the best for sharpness, lesion contours, motion artifacts, prostate border delineation, overall image quality, and diagnostic confidence. PI-RADS scoring and EPE suspicion were analyzed. Statistical methods included Wilcoxon signed-rank tests and Cohen's kappa for inter-reader agreement. Results: T2HR significantly improved lesion contours (median 5 vs. 4, p<0.001), prostate border delineation (median 5 vs. 4, p<0.001) and overall image quality (median 5 vs. 4, p<0.001) compared to T2S. However, motion artifacts were significantly worse in T2HR. Substantial inter-reader agreement was observed for PI-RADS scoring. EPE detection was marginally increased with T2HR, though histopathological validation was limited. Conclusion: T2HR imaging with DLR enhances image quality, lesion delineation, and diagnostic confidence without significant prolongation of acquisition time. It shows potential for improving EPE assessment in prostate cancer but requires further validation in larger studies.

Keywords: MRI, deep learning, Prostate, mpMRI, high-resolution

Received: 29 Aug 2025; Accepted: 06 Oct 2025.

Copyright: © 2025 Gassenmaier, Staber, Ursprung, Herrmann, Werner, Lingg, Adams, Almansour, Nikolaou and Afat. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence: Sebastian Gassenmaier, sebastian.gassenmaier@med.uni-tuebingen.de

Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.