Related Experiment Video
Updated: Jun 20, 2026
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
Diagnostic performance of biparametric versus multiparametric MRI for prostate cancer: a noninferiority, confirmatory
Xingyu Fang1, Jialin Li2, Wei Lv1
1Department of Radiology, the 305 Hospital of PLA, Beijing, China.
Objectives:
This study aimed to assess the noninferiority of biparametric MRI (bpMRI) compared to multiparametric MRI (mpMRI) using Prostate Imaging Reporting and Data System (PI-RADS) in diagnosing prostate cancer (PCa). Additionally, we compared diagnostic performance of bpMRI and mpMRI between expert and non-expert readers.
Methods:
A total of 319 men with abnormal serum prostate-specific antigen (PSA) were included in the study and underwent mpMRI examinations. Two expert readers and two non-expert readers sequentially assessed bpMRI and mpMRI using the PI-RADS scoring system, respectively. The noninferiority of bpMRI versus mpMRI was evaluated using the area under the receiver operating characteristic curve (AUROC) alongside the sensitivity and specificity at a threshold of PI-RADS 3. The diagnostic performance and inter-observer agreement of bpMRI and mpMRI were compared between expert and non-expert readers.
Results:
The prevalence of clinically significant PCa (csPCa) was 55.5% (177/319), bpMRI showed noninferior diagnostic performance to mpMRI in AUROC (0.896 [95% CI, 0.879-0.943] vs 0.911 [95% CI, 0.861-0.931]; difference of -1.5% [95% CI, -3.1% to 0.1%], p < 0.001), alongside sensitivity (92.8 [95% CI, 90.1-95.6] vs 93.8 [95% CI, 91.2-96.6]; difference of -1.0% [95% CI, -2.0% to 0.0%], p < 0.001) and specificity (69.4 [95% CI, 63.6-73.9] vs 73.1[95% CI, 67.1-76.9]; difference of -3.5% [95% CI, -6.1% to -1.4%], p < 0.01) at a threshold of PI-RADS 3. In bpMRI interpretation, specificity showed significant differences among readers (P = 0.001); expert readers (32.7%, 209/638) and non-expert readers (37.0%, 236/638) showed significant differences in PI-RADS 3 (P = 0.01) and moderate agreement (0.64 [0.53-0.71]).
Conclusion:
The diagnostic performance of bpMRI in detecting csPCa was noninferior to that of mpMRI in a selected high-risk cohort of men with abnormal serum PSA. Limited additional benefits of DCE were observed for non-expert readers. Specialized system training of PI-RADS for non-experts or diagnosis under the supervision of expert radiologists might be better alternatives.
