Purpose

In the context of analyzing neck vascular morphology, this work formulates and compares Mask R-CNN and U-Net-based algorithms to automatically segment the carotid artery (CA) and internal jugular vein (IJV) from transverse neck ultrasound (US).

Methods

US scans of the neck vasculature were collected to produce a dataset of 2439 images and their respective manual segmentations. Fourfold cross-validation was employed to train and evaluate Mask RCNN and U-Net models. The U-Net algorithm includes a post-processing step that selects the largest connected segmentation for each class. A Mask R-CNN-based vascular reconstruction pipeline was validated by performing a surface-to-surface distance comparison between US and CT reconstructions from the same patient.

Results

The average CA and IJV Dice scores produced by the Mask R-CNN across the evaluation data from all four sets were \(0.90\pm 0.08\) and \(0.88\pm 0.14\). The average Dice scores produced by the post-processed U-Net were \(0.81\pm 0.21\) and \(0.71\pm 0.23\), for the CA and IJV, respectively. The reconstruction algorithm utilizing the Mask R-CNN was capable of producing accurate 3D reconstructions with majority of US reconstruction surface points being within 2 mm of the CT equivalent.

Conclusions

On average, the Mask R-CNN produced more accurate vascular segmentations compared to U-Net. The Mask R-CNN models were used to produce 3D reconstructed vasculature with a similar accuracy to that of a manually segmented CT scan. This implementation of the Mask R-CNN network enables automatic analysis of the neck vasculature and facilitates 3D vascular reconstruction.

中文翻译：

---

从 2D 超声图像自动分割颈动脉和颈内静脉，用于 3D 血管重建。

目的

在分析颈部血管形态的背景下，这项工作制定并比较了基于 Mask R-CNN 和 U-Net 的算法，以从横向颈部超声 (US) 中自动分割颈动脉 (CA) 和颈内静脉 (IJV)。

方法

收集颈部脉管系统的 US 扫描以生成包含 2439 个图像及其各自手动分割的数据集。采用四重交叉验证来训练和评估 Mask RCNN 和 U-Net 模型。U-Net 算法包括一个后处理步骤，为每个类选择最大的连接分割。通过在来自同一患者的 US 和 CT 重建之间执行表面到表面距离比较，验证了基于 Mask R-CNN 的血管重建管道。

结果

Mask R-CNN 在所有四组评估数据中生成的平均 CA 和 IJV Dice 分数为\(0.90\pm 0.08\)和\(0.88\pm 0.14\)。对于 CA 和 IJV，后处理的 U-Net 产生的平均 Dice 分数分别为\(0.81\pm 0.21\)和\(0.71\pm 0.23\)。利用 Mask R-CNN 的重建算法能够产生准确的 3D 重建，其中大部分 US 重建表面点在 CT 等效值的 2 mm 以内。

结论

平均而言，与 U-Net 相比，Mask R-CNN 产生了更准确的血管分割。Mask R-CNN 模型用于生成 3D 重建脉管系统，其精度与手动分段 CT 扫描的精度相似。Mask R-CNN 网络的这种实现能够自动分析颈部血管系统并促进 3D 血管重建。