Department of Urology, Stanford University School of Medicine, Stanford, California.
Department of Urology, Aarhus University Hospital, Aarhus, Denmark.
J Urol. 2021 Sep;206(3):604-612. doi: 10.1097/JU.0000000000001783. Epub 2021 Apr 21.
Targeted biopsy improves prostate cancer diagnosis. Accurate prostate segmentation on magnetic resonance imaging (MRI) is critical for accurate biopsy. Manual gland segmentation is tedious and time-consuming. We sought to develop a deep learning model to rapidly and accurately segment the prostate on MRI and to implement it as part of routine magnetic resonance-ultrasound fusion biopsy in the clinic.
A total of 905 subjects underwent multiparametric MRI at 29 institutions, followed by magnetic resonance-ultrasound fusion biopsy at 1 institution. A urologic oncology expert segmented the prostate on axial T2-weighted MRI scans. We trained a deep learning model, ProGNet, on 805 cases. We retrospectively tested ProGNet on 100 independent internal and 56 external cases. We prospectively implemented ProGNet as part of the fusion biopsy procedure for 11 patients. We compared ProGNet performance to 2 deep learning networks (U-Net and holistically-nested edge detector) and radiology technicians. The Dice similarity coefficient (DSC) was used to measure overlap with expert segmentations. DSCs were compared using paired t-tests.
ProGNet (DSC=0.92) outperformed U-Net (DSC=0.85, p <0.0001), holistically-nested edge detector (DSC=0.80, p <0.0001), and radiology technicians (DSC=0.89, p <0.0001) in the retrospective internal test set. In the prospective cohort, ProGNet (DSC=0.93) outperformed radiology technicians (DSC=0.90, p <0.0001). ProGNet took just 35 seconds per case (vs 10 minutes for radiology technicians) to yield a clinically utilizable segmentation file.
This is the first study to employ a deep learning model for prostate gland segmentation for targeted biopsy in routine urological clinical practice, while reporting results and releasing the code online. Prospective and retrospective evaluations revealed increased speed and accuracy.
靶向活检可提高前列腺癌的诊断率。磁共振成像（MRI）上的精确前列腺分割对于准确活检至关重要。手动腺体分割既繁琐又耗时。我们试图开发一种深度学习模型，以便快速准确地对 MRI 上的前列腺进行分割，并将其作为临床中常规磁共振-超声融合活检的一部分实施。
共有 905 例患者在 29 个机构接受了多参数 MRI 检查，随后在 1 个机构进行了磁共振-超声融合活检。一位泌尿科肿瘤专家对轴向 T2 加权 MRI 扫描进行了前列腺分割。我们在 805 例患者上训练了一个深度学习模型 ProGNet。我们对 100 例独立的内部病例和 56 例外部病例进行了 ProGNet 的回顾性测试。我们前瞻性地将 ProGNet 作为融合活检程序的一部分应用于 11 例患者。我们将 ProGNet 的性能与 2 个深度学习网络（U-Net 和整体嵌套边缘检测器）和放射科技术员进行了比较。使用 Dice 相似系数（DSC）来衡量与专家分割的重叠程度。使用配对 t 检验比较 DSCs。
ProGNet（DSC=0.92）在回顾性内部测试集中优于 U-Net（DSC=0.85，p<0.0001）、整体嵌套边缘检测器（DSC=0.80，p<0.0001）和放射科技术员（DSC=0.89，p<0.0001）。在前瞻性队列中，ProGNet（DSC=0.93）优于放射科技术员（DSC=0.90，p<0.0001）。ProGNet 每个病例仅需 35 秒（相比之下，放射科技术员需要 10 分钟）即可生成可用于临床的分割文件。
这是第一项在常规泌尿科临床实践中针对靶向活检使用深度学习模型进行前列腺分割的研究，同时在线报告结果和发布代码。前瞻性和回顾性评估显示出更快的速度和更高的准确性。
