Comparison of Biparametric MRI and Multiparametric MRI Using Split Dynamic MRI in Prostate Cancer Diagnosis

Purpose To compare the diagnostic performance between biparametric MRI (bpMRI) and multiparametric MRI (mpMRI) using split dynamic MRI for detecting clinically significant prostate cancer (csPCa) with Prostate Imaging and Reporting Data System (PI-RADS) v2.1, using prostatectomy as the reference standard of pathologic analysis. Materials and Methods We retrospectively analyzed 77 PCa cases in 60 patients who underwent radical prostatectomy for biopsy-proven prostate cancer. The MR Images of the prostate were interpreted in consensus by two experienced radiologists (aged 3 and 35 years). We retrospectively reviewed the MRI images according to PI-RADS v2.1 and recorded the scores from T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), dynamic contrast-enhanced (DCE) imaging, and peak relaxation rate (R2*) for each lesion at three different time points: 1) bpMRI (T2WI, DWI, and ADC), 2) mpMRI1 (T2WI, DWI, ADC, and DCE), and 3) mpMRI2 (T2WI, DWI, ADC, DCE, and peak R2*). The diagnostic performances of bpMRI, mpMRI1, and mpMRI2 in detecting csPCa using PI-RADS V2.1 were evaluated. Additionally, we evaluated the diagnostic performance of DCE-MRI and peak R2* for csPCa detection. Results The area under the curve, sensitivity, specificity, and accuracy for diagnosing csPCa with PI-RADS v2.1 in all lesions and peripheral zone (PZ) lesions were not significantly different between the bpMRI, mpMRI1, and mpMRI2 protocols (p > 0.05). DCE-MRI was better than the peak R2* for diagnosing csPCa in PZ lesions (p < 0.001). Conclusion mpMRI using split dynamic MRI has diagnostic performance similar to that of bpMRI for detecting csPCa using PI-RADS v2.1.