基于U-Net的脊柱MRI骨转移瘤自动检测与分割:一项多中心研究
该多中心研究开发并评估了基于2D和3D U-Net的深度学习模型,用于脊柱MRI上骨转移瘤的自动分割与检测,使用非增强T1、增强T1 Dixon脂肪相和增强脂肪抑制T1序列的不同组合。2D U-Net T1+CE模型在外部测试中分割性能最优(Dice系数0.699,像素召回率0.653),病灶水平敏感度在内外部测试分别为0.828(497/600)和0.857(150/175),外部测试中放射科医师平均病灶敏感度为0.746、平均阳性预测值为0.701。结论认为所提出的深度学习模型在脊柱MRI骨转移瘤自动分割和检测上表现出高诊断性能。
要点
- 2D U-Net T1+CE模型在外部测试中分割性能最优,Dice系数0.699,像素召回率0.653。
- T1+CE模型病灶水平敏感度内部测试0.828(497/600),外部测试0.857(150/175)。
- 外部测试中五位放射科医师平均病灶敏感度0.746,平均病灶阳性预测值0.701。
核心结论
- 所提出的深度学习模型在脊柱MRI骨转移瘤自动分割和检测上表现出高诊断性能。
研究设计与样本
- 多中心研究,训练和内部测试纳入三个中心的302例患者共662个MRI序列(63.5±11.5岁;男:女=151:151),随机分为536和126个序列,获取时间2015年1月至2021年8月;外部测试纳入另一中心的20例患者共49个MRI序列(65.9±11.5岁;男:女=11:9),获取时间2018年1月至2020年8月。使用三个矢状位序列:非增强T1加权成像(T1)、增强T1加权Dixon脂肪相(FO)、增强脂肪抑制T1加权成像(CE);开发了7个基于2D和3D U-Net的模型,对应不同序列组合(T1、FO、CE、T1+FO、T1+CE、FO+CE、T1+FO+CE)。分割性能用Dice系数、像素召回率和像素精确率评估;检测性能用病灶水平敏感度和自由响应受试者工作特征曲线分析。外部测试集上与5位放射科医师比较。
诊断标准与关键数值
- 2D U-Net T1+CE模型在外部测试中分割性能优于其他模型,Dice系数0.699,像素召回率0.653。
- T1+CE模型病灶水平敏感度:内部测试0.828(497/600),外部测试0.857(150/175)。
- 外部测试中5位放射科医师平均病灶水平敏感度0.746,平均病灶水平阳性预测值0.701。
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
OBJECTIVE: To develop and evaluate a deep learning model for automated segmentation and detection of bone metastasis on spinal MRI.
MATERIALS AND METHODS: We included whole spine MRI scans of adult patients with bone metastasis: 662 MRI series from 302 patients (63.5 ± 11.5 years; male:female, 151:151) from three study centers obtained between January 2015 and August 2021 for training and internal testing (random split into 536 and 126 series, respectively) and 49 MRI series from 20 patients (65.9 ± 11.5 years; male:female, 11:9) from another center obtained between January 2018 and August 2020 for external testing. Three sagittal MRI sequences, including non-contrast T1-weighted image (T1), contrast-enhanced T1-weighted Dixon fat-only image (FO), and contrast-enhanced fat-suppressed T1-weighted image (CE), were used. Seven models trained using the 2D and 3D U-Nets were developed with different combinations (T1, FO, CE, T1 + FO, T1 + CE, FO + CE, and T1 + FO + CE). The segmentation performance was evaluated using Dice coefficient, pixel-wise recall, and pixel-wise precision. The detection performance was analyzed using per-lesion sensitivity and a free-response receiver operating characteristic curve. The performance of the model was compared with that of five radiologists using the external test set.
RESULTS: The 2D U-Net T1 + CE model exhibited superior segmentation performance in the external test compared to the other models, with a Dice coefficient of 0.699 and pixel-wise recall of 0.653. The T1 + CE model achieved per-lesion sensitivities of 0.828 (497/600) and 0.857 (150/175) for metastases in the internal and external tests, respectively. The radiologists demonstrated a mean per-lesion sensitivity of 0.746 and a mean per-lesion positive predictive value of 0.701 in the external test.
CONCLUSION: The deep learning models proposed for automated segmentation and detection of bone metastases on spinal MRI demonstrated high diagnostic performance.
引用
- Kim DH, Seo J, Lee JH, et al. Automated Detection and Segmentation of Bone Metastases on Spine MRI Using U-Net: A Multicenter Study. Korean J Radiol. 2024. PMID: 38528694. doi:10.3348/kjr.2023.0671
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。