文献速递肿瘤 PET/CT放射性药物与辐射防护

基于联邦深度学习的去中心化多机构协作PET衰减与散射校正

Decentralized collaborative multi-institutional PET attenuation and scatter correction using federated deep learning.
PubMed 文献 · Eur J Nucl Med Mol Imaging · 2022-12-12

本研究纳入6个中心共300例患者的18F-FDG PET图像,在数据不直接共享的前提下,用联邦学习(FL)构建PET衰减与散射校正(AC/SC)深度学习模型,并与集中式(CZ)和单中心(CB)模型比较。FL模型性能优于单中心模型、与集中式模型相当,提示联邦学习可在保护数据隐私的同时获得稳健可泛化的AC/SC模型。

要点

  • 联邦学习序列式与并行式模型的SUV绝对相对误差均与集中式模型高度一致。
  • 联邦学习模型较单中心训练策略将模型性能提升超过11%。
  • 体素级比较中联邦学习与集中式模型的决定系数相近,均明显高于单中心模型。

核心结论

  • 基于深度学习的模型在定量PET图像重建方面展现出有希望的结果。
  • 所提出的联邦学习模型性能高于单中心模型,与集中式模型相当。
  • 联邦学习框架可使AC/SC深度学习模型兼顾泛化性与稳健性,同时无需临床影像中心之间直接共享数据集。

研究设计与样本

  • 研究纳入300例患者的未校正及基于CT的衰减/散射校正(CT-ASC)18F-FDG PET图像,来自6个不同中心、每中心50例,各中心扫描仪、采集与重建方案不同。
  • 以基于CT的ASC PET图像作为标准参考,所有图像经审核纳入高质量、无伪影的PET图像,校正与未校正图像均转换为标准化摄取值(SUV)。
  • 采用改进的嵌套U-Net(U形架构中使用残差U块)。
  • 比较序列式(FL-SQ)与并行式(FL-PL)两种联邦学习模型与集中式(CZ)模型(数据汇集至一个服务器)及单中心(CB)模型的性能。
  • 各中心数据划分为训练集30例、验证集10例、测试集10例,最终评估与报告在60例患者(每中心10例)上进行。

主要发现

  • 以SUV绝对相对误差百分比(ARE%)计,FL-SQ(CI:12.21-14.81%)与FL-PL(CI:11.82-13.84%)模型均与集中式框架(CI:10.32-12.00%)高度一致。
  • 联邦学习算法较单中心训练策略(CI:22.34-26.10%)将模型性能提升超过11%。
  • 按中心分类模式下,CZ与联邦学习算法之间Mann-Whitney检验无显著差异(p值>0.05);整体数据集上不同训练方法之间观察到显著差异(p值<0.05)。
  • 相对于参考CT-ASC的体素级比较中,CZ(R2=0.94)、FL-SQ(R2=0.93)与FL-PL(R2=0.92)预测图像表现相近,而CB模型决定系数远低(R2=0.74)。
  • 尽管CZ与联邦学习方法相对参考CT-ASC呈强相关,仍观察到预测体素值存在轻度低估。

关键数值与推荐

  • 联邦学习序列式模型SUV的ARE%为CI:12.21-14.81%,并行式为CI:11.82-13.84%。
  • 集中式模型SUV的ARE%为CI:10.32-12.00%,单中心训练策略为CI:22.34-26.10%。
  • 体素级决定系数:CZ R2=0.94,FL-SQ R2=0.93,FL-PL R2=0.92,CB R2=0.74。
  • 按中心分类模式下CZ与联邦学习算法间p值>0.05,整体数据集上不同训练方法间p值<0.05。

临床怎么用

  • 原文未据此给出推荐。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

PURPOSE: Attenuation correction and scatter compensation (AC/SC) are two main steps toward quantitative PET imaging, which remain challenging in PET-only and PET/MRI systems. These can be effectively tackled via deep learning (DL) methods. However, trustworthy, and generalizable DL models commonly require well-curated, heterogeneous, and large datasets from multiple clinical centers. At the same time, owing to legal/ethical issues and privacy concerns, forming a large collective, centralized dataset poses significant challenges. In this work, we aimed to develop a DL-based model in a multicenter setting without direct sharing of data using federated learning (FL) for AC/SC of PET images. METHODS: Non-attenuation/scatter corrected and CT-based attenuation/scatter corrected (CT-ASC) 18F-FDG PET images of 300 patients were enrolled in this study. The dataset consisted of 6 different centers, each with 50 patients, with scanner, image acquisition, and reconstruction protocols varying across the centers. CT-based ASC PET images served as the standard reference. All images were reviewed to include high-quality and artifact-free PET images. Both corrected and uncorrected PET images were converted to standardized uptake values (SUVs). We used a modified nested U-Net utilizing residual U-block in a U-shape architecture. We evaluated two FL models, namely sequential (FL-SQ) and parallel (FL-PL) and compared their performance with the baseline centralized (CZ) learning model wherein the data were pooled to one server, as well as center-based (CB) models where for each center the model was built and evaluated separately. Data from each center were divided to contribute to training (30 patients), validation (10 patients), and test sets (10 patients). Final evaluations and reports were performed on 60 patients (10 patients from each center). RESULTS: In terms of percent SUV absolute relative error (ARE%), both FL-SQ (CI:12.21-14.81%) and FL-PL (CI:11.82-13.84%) models demonstrated excellent agreement with the centralized framework (CI:10.32-12.00%), while FL-based algorithms improved model performance by over 11% compared to CB training strategy (CI: 22.34-26.10%). Furthermore, the Mann-Whitney test between different strategies revealed no significant differences between CZ and FL-based algorithms (p-value > 0.05) in center-categorized mode. At the same time, a significant difference was observed between the different training approaches on the overall dataset (p-value < 0.05). In addition, voxel-wise comparison, with respect to reference CT-ASC, exhibited similar performance for images predicted by CZ (R2 = 0.94), FL-SQ (R2 = 0.93), and FL-PL (R2 = 0.92), while CB model achieved a far lower coefficient of determination (R2 = 0.74). Despite the strong correlations between CZ and FL-based methods compared to reference CT-ASC, a slight underestimation of predicted voxel values was observed. CONCLUSION: Deep learning-based models provide promising results toward quantitative PET image reconstruction. Specifically, we developed two FL models and compared their performance with center-based and centralized models. The proposed FL-based models achieved higher performance compared to center-based models, comparable with centralized models. Our work provided strong empirical evidence that the FL framework can fully benefit from the generalizability and robustness of DL models used for AC/SC in PET, while obviating the need for the direct sharing of datasets between clinical imaging centers.

来源:Eur J Nucl Med Mol Imaging,查看原文。版权归原作者所有。

引用

  1. Shiri I, Vafaei Sadr A, Akhavan A, et al. Decentralized collaborative multi-institutional PET attenuation and scatter correction using federated deep learning. Eur J Nucl Med Mol Imaging. 2022. PMID: 36508026. doi:10.1007/s00259-022-06053-8

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。