Automated 24-sector grid-map algorithm for prostate mpMRI improves precision and efficacy of prostate lesion location reporting

• Published in: European journal of radiology Vol. 183; p. 111897
• Main Authors: Walter-Rittel, Thula C., Frisch, Anne, Hamm, Charlie Alexander, Baumgärtner, Georg Lukas, Hartenstein, Alexander, Dräger, Franziska, Haas, Matthias, Cash, Hannes, Hofbauer, Sebastian, Hamm, Bernd, Beetz, Nick Lasse, Penzkofer, Tobias
• Format: Journal Article
• Language: English
• Published: Ireland Elsevier B.V 01.02.2025
• Subjects: Algorithms; Deep Learning; Humans; Image Interpretation, Computer-Assisted - methods; Magnetic Resonance Imaging; Male; Multiparametric Magnetic Resonance Imaging - methods; Peripheral Zone (Prostate); Prostate - diagnostic imaging; Prostate Segmentation; Prostatic Neoplasms - diagnostic imaging; Prostatic Neoplasms - pathology; Reproducibility of Results; Transition Zone (Prostate); Transition Zone (Prostate); CNN; MRI; Prostate Segmentation; CI; TZ; DL; PZ; FoV; Deep Learning; Magnetic Resonance Imaging; ASG; mpMRI; DWI; T2w; PIRADS; Peripheral Zone (Prostate)
• ISSN: 0720-048X; 1872-7727; 1872-7727
• DOI: 10.1016/j.ejrad.2024.111897

•A novel 24-sector automatic segmentation grid map (ASG) for prostate MRI was developed and validated.•The ASG algorithm outperformed radiologists in accurately localizing prostate lesions (80% vs 55% accuracy).•Use of the ASG-derived grid map significantly improved radiologists’ lesion localization accuracy from 55% to 71%.•The 24-sector ASG offers a practical balance between the detailed 41-sector PI-RADS v2.1 map and simpler models.•The ASG has potential to enhance reproducibility and precision in prostate MRI reporting and lesion localization. The Prostate Imaging–Reporting and Data System (PI-RADS) calls for reporting the prostate index lesion and the location within the transition (TZ) or peripheral zone (PZ) and location on a corresponding sector map. The aim of this study was to train a deep learning DL-based algorithm for automatic prostate sector mapping and to validate its’ performance. An automatic 24-sector grid-map (ASG) of the prostate was developed, based on an automatic zone-specific deep learning segmentation of the prostate. To evaluate the efficacy of the method, fiducials for random locations within the prostate were placed, and the corresponding sectors were determined for 50 mpMRI datasets. The reference standard was defined in a consensus read by two expert uroradiologists. Annotated fiducial locations were evaluated automatically by the ASG and by four radiologists in two reads with and without the help of a superimposed sector grid-map and the success rate was compared. The ASG algorithm identified the correct prostate sector of the annotated lesions in 80 % (40/50 reads) of the cases and outperformed readings of the four radiologists with 55 % (109/200), p < 0.0001. The added use of the 24 ASG map significantly improved the rate of correct sector annotation for the four radiologists to 71 % (141/200), p < 0.004. The 24 ASG map was effective for prostate sector segmentation and significantly improved location reporting of prostate lesions.