Department of Integrative Oncology, BC Cancer Research Institute, Vancouver, BC, Canada.
AI for Health, Microsoft, Redmond, Washington, USA.
Med Phys. 2024 Feb;51(2):1203-1216. doi: 10.1002/mp.16658. Epub 2023 Aug 6.
Prostate-specific membrane antigen (PSMA) PET imaging represents a valuable source of information reflecting disease stage, response rate, and treatment optimization options, particularly with PSMA radioligand therapy. Quantification of radiopharmaceutical uptake in healthy organs from PSMA images has the potential to minimize toxicity by extrapolation of the radiation dose delivery towards personalization of therapy. However, segmentation and quantification of uptake in organs requires labor-intensive organ delineations that are often not feasible in the clinic nor scalable for large clinical trials.
In this work we develop and test the PSMA Healthy organ segmentation network (PSMA-Hornet), a fully-automated deep neural net for simultaneous segmentation of 14 healthy organs representing the normal biodistribution of [ F]DCFPyL on PET/CT images. We also propose a modified U-net architecture, a self-supervised pre-training method for PET/CT images, a multi-target Dice loss, and multi-target batch balancing to effectively train PSMA-Hornet and similar networks.
The study used manually-segmented [ F]DCFPyL PET/CT images from 100 subjects, and 526 similar images without segmentations. The unsegmented images were used for self-supervised model pretraining. For supervised training, Monte-Carlo cross-validation was used to evaluate the network performance, with 85 subjects in each trial reserved for model training, 5 for validation, and 10 for testing. Image segmentation and quantification metrics were evaluated on the test folds with respect to manual segmentations by a nuclear medicine physician, and compared to inter-rater agreement. The model's segmentation performance was also evaluated on a separate set of 19 images with high tumor load.
With our best model, the lowest mean Dice coefficient on the test set was 0.826 for the sublingual gland, and the highest was 0.964 for liver. The highest mean error in tracer uptake quantification was 13.9% in the sublingual gland. Self-supervised pretraining improved training convergence, train-to-test generalization, and segmentation quality. In addition, we found that a multi-target network produced significantly higher segmentation accuracy than single-organ networks.
The developed network can be used to automatically obtain high-quality organ segmentations for PSMA image analysis tasks. It can be used to reproducibly extract imaging data, and holds promise for clinical applications such as personalized radiation dosimetry and improved radioligand therapy.
前列腺特异膜抗原（PSMA）PET 成像提供了有价值的信息源，反映了疾病的分期、反应率和治疗优化选择，特别是使用 PSMA 放射性配体治疗。从 PSMA 图像中量化放射性药物在健康器官中的摄取量，通过推断放射性剂量的传递，可以潜在地最小化毒性，实现治疗的个体化。然而，对器官的摄取进行分割和定量需要进行劳动密集型的器官描绘，而在临床上通常是不可行的，也不适合大型临床试验。
在这项工作中，我们开发并测试了 PSMA 健康器官分割网络（PSMA-Hornet），这是一种用于同时分割 14 个健康器官的全自动深度神经网络，代表了[F]DCFPyL 在 PET/CT 图像上的正常生物分布。我们还提出了一种修改后的 U 型网络架构、一种用于 PET/CT 图像的自我监督预训练方法、一种多目标 Dice 损失和多目标批量平衡，以有效地训练 PSMA-Hornet 和类似的网络。
该研究使用了 100 名患者的手动分割[F]DCFPyL PET/CT 图像和 526 张无分割图像。未分割的图像用于自我监督模型的预训练。对于有监督的训练，采用蒙特卡罗交叉验证来评估网络性能，每次试验中保留 85 名患者用于模型训练，5 名用于验证，10 名用于测试。使用核医学医师的手动分割来评估测试折叠的图像分割和量化指标，并与组内一致性进行比较。还在一个具有高肿瘤负荷的 19 张图像的独立数据集上评估了模型的分割性能。
在我们的最佳模型中，舌下腺的最低平均 Dice 系数为 0.826，肝脏的最高为 0.964。在示踪剂摄取定量中，最高的平均误差为 13.9%，发生在舌下腺。自我监督的预训练提高了训练收敛性、训练-测试泛化和分割质量。此外，我们发现多目标网络产生的分割精度明显高于单器官网络。
所开发的网络可用于自动获取 PSMA 图像分析任务的高质量器官分割。它可用于重复提取成像数据，并有望在个性化放射剂量学和改良放射性配体治疗等临床应用中得到应用。
