Related Experiment Video
Updated: Oct 17, 2025
A Cognitive Fusion-guided Prostate Biopsy Using Multiparametric Magnetic Resonance Imaging and Transrectal Ultrasound
Published on: March 21, 2025
Synthesizing High-b-Value Diffusion-weighted Imaging of the Prostate Using Generative Adversarial Networks
Lei Hu1, Da-Wei Zhou1, Yun-Fei Zha1
1Department of Diagnostic and Interventional Radiology, Shanghai Jiao Tong University Affiliated Sixth People's Hospital, 600 Yi Shan Road, Shanghai 200233, China (L.H., W.H.X., J.G.Z.); State Key Laboratory of Integrated Services Networks, School of Telecommunications Engineering, Xidian University, Xi'an, China (D.W.Z.); Department of Radiology, Renmin Hospital, Wuhan University, Wuhan, China (Y.F.Z., L.L., H. He, L.Q., Y.K.Z.); MR Application Development, Siemens Shenzhen MR, Shenzhen, China (C.X.F.); and Department of Radiology, The Affiliated Renmin Hospital of Jiangsu University, Zhenjiang, China (H. Hu).
Purpose:
To develop and evaluate a diffusion-weighted imaging (DWI) deep learning framework based on the generative adversarial network (GAN) to generate synthetic high-b-value (b =1500 sec/mm2) DWI (SYNb1500) sets from acquired standard-b-value (b = 800 sec/mm2) DWI (ACQb800) and acquired standard-b-value (b = 1000 sec/mm2) DWI (ACQb1000) sets.
Materials And Methods:
This retrospective multicenter study included 395 patients who underwent prostate multiparametric MRI. This cohort was split into internal training (96 patients) and external testing (299 patients) datasets. To create SYNb1500 sets from ACQb800 and ACQb1000 sets, a deep learning model based on GAN (M0) was developed by using the internal dataset. M0 was trained and compared with a conventional model based on the cycle GAN (Mcyc). M0 was further optimized by using denoising and edge-enhancement techniques (optimized version of the M0 [Opt-M0]). The SYNb1500 sets were synthesized by using the M0 and the Opt-M0 were synthesized by using ACQb800 and ACQb1000 sets from the external testing dataset. For comparison, traditional calculated (b =1500 sec/mm2) DWI (CALb1500) sets were also obtained. Reader ratings for image quality and prostate cancer detection were performed on the acquired high-b-value (b = 1500 sec/mm2) DWI (ACQb1500), CALb1500, and SYNb1500 sets and the SYNb1500 set generated by the Opt-M0 (Opt-SYNb1500). Wilcoxon signed rank tests were used to compare the readers' scores. A multiple-reader multiple-case receiver operating characteristic curve was used to compare the diagnostic utility of each DWI set.
Results:
When compared with the Mcyc, the M0 yielded a lower mean squared difference and higher mean scores for the peak signal-to-noise ratio, structural similarity, and feature similarity (P < .001 for all). Opt-SYNb1500 resulted in significantly better image quality (P ≤ .001 for all) and a higher mean area under the curve than ACQb1500 and CALb1500 (P ≤ .042 for all).
Conclusion:
A deep learning framework based on GAN is a promising method to synthesize realistic high-b-value DWI sets with good image quality and accuracy in prostate cancer detection.Keywords: Prostate Cancer, Abdomen/GI, Diffusion-weighted Imaging, Deep Learning Framework, High b Value, Generative Adversarial Networks© RSNA, 2021 Supplemental material is available for this article.
