文献速递神经MR

门控高分辨率神经网络的构建与评估:用于脑转移瘤自动检测和分割

Construction and evaluation of a gated high-resolution neural network for automatic brain metastasis detection and segmentation.
PubMed 文献 · Eur Radiol · 2023-04-25

这项回顾性研究纳入1392例有脑转移瘤(共14542个)和200例无脑转移瘤患者的颅脑MRI,构建门控高分辨率卷积神经网络(GHR-CNN)用于检测和分割脑转移瘤,并在其他医院的独立数据集上检验泛化性。模型在测试集recall 0.88、DSC 0.90、PPV 0.93、FP 1.01,外部数据集recall 0.85、DSC 0.89、PPV 0.93、FP 1.07;模型辅助使4位放射科医生的脑转移瘤检出率提高5.2%–15.1%(均p<0.001),≤5 mm小病灶检出率提高7.2%–27.0%(均p<0.001)。

要点

  • GHR-CNN可在增强3D-T1W图像上准确检测和分割脑转移瘤。
  • 模型辅助使4位放射科医生脑转移瘤检出率提高5.2%–15.1%(均p<0.001)。
  • 模型辅助使≤5 mm小转移瘤检出率提高7.2%–27.0%(均p<0.001)。

核心结论

  • 所提出的模型能准确检测和分割脑转移瘤,具有更高的敏感度和更少的时间消耗,显示出提升放射科医生脑转移瘤检出表现的潜力。

研究设计与样本

  • 通过比较模型与放射科医生的脑转移瘤检测和分割表现,以及比较放射科医生在有和无模型辅助下的病灶检出表现,评估模型的临床使用潜力。

关键影像表现/征象

  • GHR-CNN可在增强3D-T1W图像上准确检测和分割脑转移瘤。
  • 模型在测试集recall 0.88、DSC 0.90、PPV 0.93、FP 1.01。
  • 模型在其他医院数据集recall 0.85、DSC 0.89、PPV 0.93、FP 1.07。

诊断标准与关键数值

  • 测试集:recall 0.88,DSC 0.90,PPV 0.93,FP 1.01。
  • 其他医院数据集:recall 0.85,DSC 0.89,PPV 0.93,FP 1.07。
  • 模型辅助下4位放射科医生脑转移瘤检出率提高5.2%–15.1%(均p<0.001)。
  • 模型辅助下≤5 mm小脑转移瘤检出率提高7.2%–27.0%(均p<0.001)。

临床应用与阅片要点

  • 本研究提供了一种有前景的计算机辅助工具,用于在常规临床实践中辅助癌症患者的脑转移瘤检测和分割。
  • GHR-CNN可自动分割脑转移瘤,且时间很短。

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)。

英文原文(摘要)· 点击展开

OBJECTIVES: To construct and evaluate a gated high-resolution convolutional neural network for detecting and segmenting brain metastasis (BM).

METHODS: This retrospective study included craniocerebral MRI scans of 1392 patients with 14,542 BMs and 200 patients with no BM between January 2012 and April 2022. A primary dataset including 1000 cases with 11,686 BMs was employed to construct the model, while an independent dataset including 100 cases with 1069 BMs from other hospitals was used to examine the generalizability. The potential of the model for clinical use was also evaluated by comparing its performance in BM detection and segmentation to that of radiologists, and comparing radiologists' lesion detecting performances with and without model assistance.

RESULTS: Our model yielded a recall of 0.88, a dice similarity coefficient (DSC) of 0.90, a positive predictive value (PPV) of 0.93 and a false positives per patient (FP) of 1.01 in the test set, and a recall of 0.85, a DSC of 0.89, a PPV of 0.93, and a FP of 1.07 in dataset from other hospitals. With the model's assistance, the BM detection rates of 4 radiologists improved significantly, ranging from 5.2 to 15.1% (all p < 0.001), and also for detecting small BMs with diameter ≤ 5 mm (ranging from 7.2 to 27.0%, all p < 0.001).

CONCLUSIONS: The proposed model enables accurate BM detection and segmentation with higher sensitivity and less time consumption, showing the potential to augment radiologists' performance in detecting BM.

CLINICAL RELEVANCE STATEMENT: This study offers a promising computer-aided tool to assist the brain metastasis detection and segmentation in routine clinical practice for cancer patients.

KEY POINTS: • The GHR-CNN could accurately detect and segment BM on contrast-enhanced 3D-T1W images. • The GHR-CNN improved the BM detection rate of radiologists, including the detection of small lesions. • The GHR-CNN enabled automated segmentation of BM in a very short time.

来源:Eur Radiol,查看原文。版权归原作者所有。

引用

  1. Qu J, Zhang W, Shu X, et al. Construction and evaluation of a gated high-resolution neural network for automatic brain metastasis detection and segmentation. Eur Radiol. 2023. PMID: 37186214. doi:10.1007/s00330-023-09648-3

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。