Related Experiment Video
Updated: Sep 14, 2026
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
Apparent diffusion coefficient-based single-sequence radiomics integrated with clinical variables and Prostate
Jiong Huang1, Yulei Wan1, Junyan Wang1
1Department of Radiology, The Sixth Hospital of Wuhan, Affiliated Hospital of Jianghan University, Wuhan, China.
Background:
Accurate identification of clinically significant prostate cancer (csPCa) is essential for reducing unnecessary biopsy and overtreatment. This study aimed to develop and internally validate an apparent diffusion coefficient (ADC)-based single-sequence radiomics model integrated with clinical variables and Prostate Imaging Reporting and Data System (PI-RADS) for predicting csPCa.
Methods:
This retrospective single-center study included 301 patients who underwent prostate magnetic resonance imaging (MRI) and histopathological assessment between January 2020 and February 2026. Patients were stratified into training and test cohorts at a 7:3 ratio. Radiomics features were extracted from manually delineated ADC-based index lesions, and a parsimonious radiomics score (Rad-score) was developed using reproducibility filtering, redundancy reduction, one-standard-error least absolute shrinkage and selection operator (LASSO) selection, and stability- and sample-size-based complexity control. Full radiomics-pipeline repeated nested cross-validation was performed for internal validation. Five logistic regression models were developed: clinical, PI-RADS, ADC radiomics, clinicoradiological, and combined models. Model performance was evaluated using discrimination, calibration, and decision curve analysis.
Results:
The cohort included 82 patients with csPCa and 219 with non-csPCa. Five ADC radiomics features were retained for Rad-score construction. In the test cohort, the ADC radiomics model achieved an area under the receiver operating characteristic curve (AUC) of 0.916. The clinicoradiological and combined models achieved AUCs of 0.944 and 0.951, respectively, with no significant improvement after adding the Rad-score (ΔAUC=0.007; P=0.71).
Conclusions:
ADC radiomics showed strong standalone discrimination for csPCa, but its addition to clinical variables and PI-RADS did not significantly improve overall discrimination. These findings support transparent evaluation of radiomics against established clinicoradiological assessment but require independent external validation before clinical implementation.
