Preoperative apical margin risk estimation from T2-weighted MRI radiomics in prostate cancer with MRI-visible apical tumors: toward multidisciplinary decision support.

Authors

Yi Liu, Yi-Ze Zhao, Shui Liu, Xian-Shu Gao, Ming-Wei Ma

Published in

Asian journal of andrology. Oct 20, 2026. Epub Oct 20, 2026.

Abstract

This retrospective study developed and validated a T2-weighted imaging (T2WI)-based radiomics model for preoperative prediction of apical positive surgical margins (PSMs) in patients undergoing radical prostatectomy for prostate cancer. Overall, 235 patients (74 with apical PSMs and 161 without apical PSMs) were randomly divided into a training set (n = 164; 69.8%) and an independent test set (n = 71; 30.2%). Eligible patients had preoperative multiparametric magnetic resonance imaging within 6 weeks before surgery, an index lesion involving or abutting the prostatic apex, definitive apical margin status, and standard radical prostatectomy without nerve-sparing. Patients with inadequate imaging for apical assessment or region-of-interest delineation, prior transurethral resection of the prostate, or neoadjuvant androgen-deprivation therapy were excluded. A total of 1409 radiomic features were extracted from T2WI. After normalization and removal of highly correlated features (|r| > 0.99), six optimal features were selected by recursive feature elimination. A gradient boosting machine model was constructed, and performance was assessed using the area under the receiver operating characteristic curve (AUC), Brier score, and decision curve analysis. The apical PSM rate was 31.5%. In the test set, the model achieved an AUC of 0.874, Brier score of 0.120, accuracy of 87.3%, sensitivity of 81.8%, and specificity of 89.8%. Skewness was the most important feature, contributing 34.8% to the model. Apical PSMs were significantly associated with extraprostatic extension (P = 0.001), and the tumor-margin model provided the greatest net clinical benefit. T2WI-based radiomics may support individualized preoperative risk assessment and surgical planning for prostate cancer.

PMID:
42902531
Bibliographic data and abstract were imported from PubMed on 21 Oct 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.