文献速递肿瘤 PET/CT

使用TriDFusion(3DF)的棕色脂肪组织机器学习nnU-Net V2网络

Brown adipose tissue machine learning nnU-Net V2 network using TriDFusion (3DF).
PubMed 文献 · EJNMMI Res · 2025-08-13

本研究利用基于阈值半自动的3DF工作流程分割317例含棕色脂肪组织的PET/CT扫描,经医师审核修正后训练nnU-Net V2模型,并应用于118例独立PET/CT扫描,医师评估网络自动分割在各解剖部位准确率约90%。作者认为nnU-Net V2能有效识别棕色脂肪组织,但实现完美分割仍具挑战。

要点

  • 采用半自动阈值3DF工作流程分割317例含棕色脂肪组织的PET/CT扫描,经三位医师审核修正后用于训练nnU-Net V2模型。
  • 将训练好的模型应用于118例独立PET/CT扫描,医师评估自动分割在各解剖部位准确率约90%。
  • 作者认为nnU-Net V2能有效识别棕色脂肪组织,但受PET/CT配准不良及连续层面无可见棕色脂肪活性等因素影响,完美分割仍具挑战。

核心结论

  • 尽管nnU-Net V2能有效从PET/CT扫描中识别棕色脂肪组织,但训练出能完美分割棕色脂肪组织的AI模型仍具挑战。

研究设计与样本

  • 本研究使用半自动、基于阈值的3DF工作流程分割317例含棕色脂肪组织的PET/CT扫描,经三位医师审核修正后训练nnU-Net V2模型,并应用于118例独立PET/CT扫描。

主要发现

  • 医师审核网络对118例独立PET/CT扫描的自动分割,评估各解剖部位准确率约90%。

局限性

  • 训练完美棕色脂肪组织分割AI模型受PET/CT配准不良及连续层面无可见棕色脂肪活性等因素影响。

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

BACKGROUND: Recent advances in machine learning have revolutionized medical imaging. Currently, identifying brown adipose tissue (BAT) relies on manual identification and segmentation on Fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) scans. However, the process is time-consuming, especially for studies involving a large number of cases, and is subject to bias due to observer dependency. The introduction of machine learning algorithms, such as the PET/CT algorithm implemented in the TriDFusion (3DF) Image Viewer, represents a significant advancement in BAT detection. In the context of cancer care, artificial intelligence (AI)-driven BAT detection holds immense promise for rapid and automatic differentiation between malignant lesions and non-malignant BAT confounds. By leveraging machine learning to discern intricate patterns in imaging data, this study aims to advance the automation of BAT recognition and provide precise quantitative assessment of radiographic features. RESULTS: We used a semi-automatic, threshold-based 3DF workflow to segment 317 PET/CT scans containing BAT. To minimize manual edits, we defined exclusion zones via machine-learning-based CT organ segmentation and used those organ masks to assign each volume of interest (VOI) to its anatomical site. Three physicians then reviewed and corrected all segmentations using the 3DF contour panel. The final, edited masks were used to train an nnU-Net V2 model, which we subsequently applied to 118 independent PET/CT scans. Across all anatomical sites, physicians reviewed the network’s automated segmentations to be approximately 90% accurate. CONCLUSION: Although nnU-Net V2 effectively identified BAT from PET/CT scans, training an AI model capable of perfect BAT segmentation remains a challenge due to factors such as PET/CT misregistration and the absence of visible BAT activity across contiguous slices.

来源:EJNMMI Res,查看原文。版权归原作者所有。

引用

  1. Lafontaine D, Chahwan S, Barraza G, et al. Brown adipose tissue machine learning nnU-Net V2 network using TriDFusion (3DF). EJNMMI Res. 2025. PMID: 40804586. doi:10.1186/s13550-025-01303-w

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。