Keywords: Prostate, Prostate, Segmentation
Motivation: Measuring membranous urethral length (MUL) on MRI can predict urinary continence outcomes after prostate cancer surgery, but requires expert radiologists - a resource many hospitals lack. Manual measurements are time-consuming and subjective, creating treatment planning delays.
Goal(s): Develop an automated deep learning system to identify, segment, and measure MUL from prostate MRI scans with accuracy comparable to expert radiologists.
Approach: Created an AI pipeline using a modified U-Net architecture: optimal slice selection, MU segmentation via bounding box prediction, and automated MUL calculation.
Results: System achieved sub-millimeter accuracy (mean difference 0.9±1.5mm) from expert measurements, with consistent performance across anatomical variations and validation by senior radiologists.
Impact: By eliminating the need for specialized radiologist expertise, this automated system could enable widespread adoption of MUL-based surgical planning in resource-limited settings, helping surgeons optimize their approach to preserve urinary continence for prostate cancer patients.
How to access this content:
For one year after publication, abstracts and videos are only open to registrants of this annual meeting. Registrants should use their existing login information. Non-registrant access can be purchased via the ISMRM E-Library.
After one year, current ISMRM & ISMRT members get free access to both the abstracts and videos. Non-members and non-registrants must purchase access via the ISMRM E-Library.
After two years, the meeting proceedings (abstracts) are opened to the public and require no login information. Videos remain behind password for access by members, registrants and E-Library customers.
Keywords