Department of Medical Physics, Memorial Sloan Kettering Cancer, Center, New York, New York, USA.
Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, New York, USA.
Med Phys. 2023 Aug;50(8):4854-4870. doi: 10.1002/mp.16320. Epub 2023 Mar 13.
Dose escalation radiotherapy enables increased control of prostate cancer (PCa) but requires segmentation of dominant index lesions (DIL). This motivates the development of automated methods for fast, accurate, and consistent segmentation of PCa DIL.
To construct and validate a model for deep-learning-based automatic segmentation of PCa DIL defined by Gleason score (GS) ≥3+4 from MR images applied to MR-guided radiation therapy. Validate generalizability of constructed models across scanner and acquisition differences.
Five deep-learning networks were evaluated on apparent diffusion coefficient (ADC) MRI from 500 lesions in 365 patients arising from internal training Dataset 1 (156 lesions in 125 patients, 1.5Tesla GE MR with endorectal coil), testing using Dataset 1 (35 lesions in 26 patients), external ProstateX Dataset 2 (299 lesions in 204 patients, 3Tesla Siemens MR), and internal inter-rater Dataset 3 (10 lesions in 10 patients, 3Tesla Philips MR). The five networks include: multiple resolution residually connected network (MRRN) and MRRN regularized in training with deep supervision implemented into the last convolutional block (MRRN-DS), Unet, Unet++, ResUnet, and fast panoptic segmentation (FPSnet) as well as fast panoptic segmentation with smoothed labels (FPSnet-SL). Models were evaluated by volumetric DIL segmentation accuracy using Dice similarity coefficient (DSC) and the balanced F1 measure of detection accuracy, as a function of lesion aggressiveness and size (Dataset 1 and 2), and accuracy with respect to two-raters (on Dataset 3). Upon acceptance for publication segmentation models will be made available in an open-source GitHub repository.
In general, MRRN-DS more accurately segmented tumors than other methods on the testing datasets. MRRN-DS significantly outperformed ResUnet in Dataset2 (DSC of 0.54 vs. 0.44, p < 0.001) and the Unet++ in Dataset3 (DSC of 0.45 vs. p = 0.04). FPSnet-SL was similarly accurate as MRRN-DS in Dataset2 (p = 0.30), but MRRN-DS significantly outperformed FPSnet and FPSnet-SL in both Dataset1 (0.60 vs. 0.51 [p = 0.01] and 0.54 [p = 0.049] respectively) and Dataset3 (0.45 vs. 0.06 [p = 0.002] and 0.24 [p = 0.004] respectively). Finally, MRRN-DS produced slightly higher agreement with experienced radiologist than two radiologists in Dataset 3 (DSC of 0.45 vs. 0.41).
MRRN-DS was generalizable to different MR testing datasets acquired using different scanners. It produced slightly higher agreement with an experienced radiologist than that between two radiologists. Finally, MRRN-DS more accurately segmented aggressive lesions, which are generally candidates for radiative dose ablation.
剂量递增放疗使前列腺癌（PCa）的控制得到了提高，但需要对优势指数病变（DIL）进行分割。这促使人们开发了用于快速、准确和一致地分割 PCa DIL 的自动化方法。
构建并验证一种基于深度学习的模型，用于从磁共振图像中对 Gleason 评分（GS）≥3+4 的 PCa DIL 进行自动分割，该模型适用于磁共振引导放疗。验证构建的模型在扫描仪和采集差异方面的通用性。
在来自 365 名患者的 500 个病变的表观扩散系数（ADC）MRI 上评估了 5 种深度学习网络，这些病变来自内部训练数据集 1（125 名患者中的 156 个病变，1.5Tesla GE MR 带有直肠内线圈），使用数据集 1（26 名患者中的 35 个病变）、外部 ProstateX 数据集 2（204 名患者中的 299 个病变，3Tesla 西门子 MR）和内部观察者间数据集 3（10 名患者中的 10 个病变，3Tesla Philips MR）进行测试。这 5 种网络包括：多分辨率残差连接网络（MRRN）和在训练中用深度监督正则化的 MRRN（MRRN-DS）、Unet、Unet++、ResUnet 和快速全景分割（FPSnet）以及带有平滑标签的快速全景分割（FPSnet-SL）。通过使用 Dice 相似系数（DSC）和检测准确性的平衡 F1 度量来评估病变体积分割准确性，评估模型的功能是根据病变侵袭性和大小（数据集 1 和 2），以及与两位观察者的准确性（数据集 3）。一旦被接受发表，分割模型将在一个开源 GitHub 存储库中提供。
一般来说，MRRN-DS 在测试数据集上比其他方法更准确地分割了肿瘤。MRRN-DS 在数据集 2 中显著优于 ResUnet（DSC 为 0.54 与 0.44，p<0.001）和数据集 3 中的 Unet++（DSC 为 0.45 与 p=0.04）。FPSnet-SL 在数据集 2 中的准确性与 MRRN-DS 相似（p=0.30），但 MRRN-DS 在数据集 1（0.60 与 0.51 [p=0.01] 和 0.54 [p=0.049]）和数据集 3（0.45 与 0.06 [p=0.002] 和 0.24 [p=0.004]）中均显著优于 FPSnet 和 FPSnet-SL。最后，MRRN-DS 在数据集 3 中与经验丰富的放射科医生的一致性略高于两位放射科医生（DSC 为 0.45 与 0.41）。
MRRN-DS 可推广到使用不同扫描仪采集的不同磁共振测试数据集。它与一位经验丰富的放射科医生的一致性略高于两位放射科医生之间的一致性。最后，MRRN-DS 更准确地分割了侵袭性病变，这些病变通常是放射性剂量消融的候选病变。
