Department of Automation, School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China.
Shandong Cancer Hospital Affiliated to Shandong University, Jinan, 250117, China.
Eur Radiol. 2023 May;33(5):3521-3531. doi: 10.1007/s00330-023-09420-7. Epub 2023 Jan 25.
To develop and validate a two-stage deep learning model for automatic detection and segmentation of brain metastases (BMs) in MRI images.
In this retrospective study, T1-weighted (T1) and T1-weighted contrast-enhanced (T1ce) MRI images of 649 patients who underwent radiotherapy from August 2019 to January 2022 were included. A total of 5163 metastases were manually annotated by neuroradiologists. A two-stage deep learning model was developed for automatic detection and segmentation of BMs, which consisted of a lightweight segmentation network for generating metastases proposals and a multi-scale classification network for false-positive suppression. Its performance was evaluated by sensitivity, precision, F1-score, dice, and relative volume difference (RVD).
Six hundred forty-nine patients were randomly divided into training (n = 295), validation (n = 99), and testing (n = 255) sets. The proposed two-stage model achieved a sensitivity of 90% (1463/1632) and a precision of 56% (1463/2629) on the testing set, outperforming one-stage methods based on a single-shot detector, 3D U-Net, and nnU-Net, whose sensitivities were 78% (1276/1632), 79% (1290/1632), and 87% (1426/1632), and the precisions were 40% (1276/3222), 51% (1290/2507), and 53% (1426/2688), respectively. Particularly for BMs smaller than 5 mm, the proposed model achieved a sensitivity of 66% (116/177), far superior to one-stage models (21% (37/177), 36% (64/177), and 53% (93/177)). Furthermore, it also achieved high segmentation performance with an average dice of 81% and an average RVD of 20%.
A two-stage deep learning model can detect and segment BMs with high sensitivity and low volume error.
• A two-stage deep learning model based on triple-channel MRI images identified brain metastases with 90% sensitivity and 56% precision. • For brain metastases smaller than 5 mm, the proposed two-stage model achieved 66% sensitivity and 22% precision. • For segmentation of brain metastases, the proposed two-stage model achieved a dice of 81% and a relative volume difference (RVD) of 20%.
开发并验证一种用于 MRI 图像中脑转移瘤（BMs）自动检测和分割的两阶段深度学习模型。
本回顾性研究纳入了 2019 年 8 月至 2022 年 1 月期间接受放疗的 649 名患者的 T1 加权（T1）和 T1 增强对比（T1ce）MRI 图像。共有神经放射科医生手动标注了 5163 个转移瘤。我们开发了一种用于 BMs 自动检测和分割的两阶段深度学习模型，它由一个轻量级分割网络生成转移瘤提案和一个多尺度分类网络用于抑制假阳性。通过灵敏度、精度、F1 评分、Dice 和相对体积差异（RVD）来评估其性能。
649 名患者被随机分为训练集（n=295）、验证集（n=99）和测试集（n=255）。所提出的两阶段模型在测试集上的灵敏度为 90%（1463/1632），精度为 56%（1463/2629），优于基于单次检测、3D U-Net 和 nnU-Net 的单阶段方法，其灵敏度分别为 78%（1276/1632）、79%（1290/1632）和 87%（1426/1632），精度分别为 40%（1276/3222）、51%（1290/2507）和 53%（1426/2688）。特别是对于小于 5mm 的 BMs，该模型的灵敏度达到了 66%（116/177），远高于单阶段模型（21%（37/177）、36%（64/177）和 53%（93/177））。此外，它还具有较高的分割性能，平均 Dice 为 81%，平均 RVD 为 20%。
两阶段深度学习模型可以实现高灵敏度和低体积误差的脑转移瘤检测和分割。
基于三通道 MRI 图像的两阶段深度学习模型识别脑转移瘤的灵敏度为 90%，精度为 56%。
对于小于 5mm 的脑转移瘤，所提出的两阶段模型的灵敏度达到 66%，精度为 22%。
对于脑转移瘤的分割，所提出的两阶段模型的 Dice 为 81%，RVD 为 20%。
