MR Virtual Cytometry for Zone-specific Prostate Cancer Characterization with Histologic and Molecular Validation

Purpose To evaluate time-dependent diffusion MRI (t_d-dMRI)-based virtual cytometry for characterizing prostatic lesions and assess its zone-specific diagnostic value. Materials and Methods In this prospective study, participants with clinical suspicion of PCa were enrolled between March 2023 and August 2024 for t_d-dMRI. By modeling prostate tissue as intra- and extracellular compartments, microstructural features, including intracellular volume fraction, cell diameter, cellularity, and transmembrane water exchange time, were estimated. Diagnostic performance was analyzed using a generalized linear mixed model to account for multiple lesions per participant. Validation included correlations with whole-slide histologic examination as the reference standard (n = 36) and prostate-specific membrane antigen (PSMA) PET (n = 27). Results A total of 117 male participants (mean age, 70.0 years ± 6.8 [SD]) contributing 199 lesions were analyzed. PCa showed elevated intracellular volume fraction, cellularity, and accelerated transmembrane water exchange compared with benign lesions (adjusted P < .05). Higher Gleason grades were associated with further increases in these parameters (odds ratio, 2.62, 2.43, and 0.62, respectively; P < .05). Zone-specific features included higher intracellular water diffusivity in the peripheral zone and smaller cell size in the transition zone (TZ) (all adjusted P < .05). The t_d-dMRI-derived cellularity index demonstrated the highest performance for discriminating PCa from benign lesions, particularly in the TZ (area under the receiver operating characteristic curve, 0.94), which improved to 0.96 when combined with the Prostate Imaging Reporting and Data System. Validation showed strong correlations with histopathologic examination (cancer only r = 0.81, P < .001) and PSMA PET (r = 0.74, P < .05). Conclusion In this study, t_d-dMRI-based MR virtual cytometry enabled quantitative characterization of cellular microstructure in prostate disease and improved diagnostic performance, particularly in the TZ. Keywords: MR-Diffusion-weighted Imaging, PET, Prostate, Tissue Characterization, Modeling Supplemental material is available for this article. © RSNA, 2026.