基于多尺度聚合与高分辨率增强的CTA脑血管分割模型

doi:10.19678/j.issn.1000-3428.0070221

摘要/Abstract

摘要：

在颅脑CT血管造影(CTA)图像中, 脑血管形态各异、分布分散且不同患者之间差异较大。这导致利用U-Net进行血管分割时对血管局部形态的适应性不足, 容易忽略分散目标之间的相关性, 且在下采样过程中会丢失小尺度血管信息。针对以上问题, 在U-Net的基础上进行改进, 提出一种基于多尺度聚合和高分辨率增强的血管分割网络BVU-Net。在编码器的瓶颈层设计一种结合空洞变形金字塔(DDP)路径与全局注意力(GA)路径的多尺度特征聚合(MSFA)模块, 旨在同时捕获血管的不同尺度的局部形态特征和全局空间相关性特征。在跳跃连接路径中设计高分辨率特征增强(HRFE)模块, 使模型能充分利用语义信息更丰富的高级特征, 提高浅层高分辨率特征的表征能力, 补充小血管信息, 进一步提升血管分割精度。BVU-Net模型在公开数据集3D-IRCADb和私有数据集GLCTA上进行实验验证, Dice指标分别达到0.787 2和0.924 8, 平均交并比(MIoU)指标分别达到0.832 2和0.932 1。上述结果表明, BVU-Net模型的表现优于其他基于U-Net的改进分割模型, 具有一定泛化能力, 为后续的临床治疗和预后分析提供了更有力的参考。

关键词: 脑血管分割, 急性缺血性卒中, 多尺度特征聚合, 高分辨率增强, 可变形卷积

Abstract:

Cerebral vessels in brain CT Angiography (CTA) images exhibit diverse morphologies and distributions with significant variations among patients. The standard U-Net often struggles to adapt to local vessel morphology, leading to the loss of small target information during down-sampling and neglecting the correlations among scattered objects. To address these challenges, this study enhances the U-Net architecture and introduces the BVU-Net, a cerebral vessel segmentation network that utilizes multi-scale aggregation and high-resolution enhancement. The BVU-Net designs a Multi-Scale Feature Aggregation (MSFA) module in its bottleneck layer, which captures local vessel features at various scales as well as global correlation features. This module integrates the Dilated Deformable Pyramid (DDP) path and the Global Attention (GA) path. In addition, a High-Resolution Feature Enhancement (HRFE) module is incorporated into the skip connection paths, allowing for the effective use of advanced features with richer semantic information. This enhancement improves the representation of high-resolution features and supplements the information on small vessels. The performance of the BVU-Net is evaluated on the public dataset 3D-IRCADb and the private dataset GLCTA, achieving Dice scores of 0.787 2 and 0.924 8 and Mean Intersection over Union (MIoU) scores of 0.832 2 and 0.932 1, respectively. These results demonstrate that the BVU-Net outperforms other improved U-Net segmentation models and exhibits notable generalization capabilities, providing valuable insights for future clinical treatment and prognosis analysis.

Key words: cerebral vessel segmentation, Acute Ischemic Stroke (AIS), Multi-Scale Feature Aggregation (MSFA), high-resolution enhancement, deformable convolution

张天旭, 黄慧, 黄丙仓, 马燕, 徐傲, 李晓艳, 周孝雯, 刘之之. 基于多尺度聚合与高分辨率增强的CTA脑血管分割模型[J]. 计算机工程, 2025, 51(4): 37-46.

ZHANG Tianxu, HUANG Hui, HUANG Bingcang, MA Yan, XU Ao, LI Xiaoyan, ZHOU Xiaowen, LIU Zhizhi. CTA Cerebral Vessel Segmentation Model Based on Multi-scale Aggregation and High-resolution Enhancement[J]. Computer Engineering, 2025, 51(4): 37-46.

https://www.ecice06.com/CN/Y2025/V51/I4/37

图/表 11

图1 BVU-Net结构

Fig.1 BVU-Net structure

图2 MSFA模块结构

Fig.2 MSFA module structure

图3 HRFE模块结构

Fig.3 HRFE module structure

图4 3D-IRCADb数据集标注示例

Fig.4 Annotation example of 3D-IRCADb dataset

图5 GLCTA数据集标注示例

Fig.5 Annotation example of GLCTA dataset

图6 3D-IRCADb数据集上的分割结果可视化

Fig.6 Visualization of segmentation results on the 3D-IRCADb dataset

图7 GLCTA数据集上的分割结果可视化

Fig.7 Visualization of segmentation results on the GLCTA dataset

参考文献 32

1	王陇德, 彭斌, 张鸿祺, 等. 《中国脑卒中防治报告2020》概要. 中国脑血管病杂志, 2022, 19 (2): 136- 144. doi: 10.3969/j.issn.1672-5921.2022.02.011
	WANG L D , PENG B , ZHANG H Q , et al. Brief report on stroke prevention and treatment in China, 2020. Chinese Journal of Cerebrovascular Diseases, 2022, 19 (2): 136- 144. doi: 10.3969/j.issn.1672-5921.2022.02.011
2	霍晓川, 高峰. 中国急性缺血性卒中血管内治疗现状初步调查. 中国卒中杂志, 2016, 11 (4): 283- 287. doi: 10.3969/j.issn.1673-5765.2016.04.008
	HUO X C , GAO F . Preliminary investigation of endovascular treatment status of acute ischemic stroke in China. Chinese Journal of Stroke, 2016, 11 (4): 283- 287. doi: 10.3969/j.issn.1673-5765.2016.04.008
3	SOARES J V , LEANDRO J J , JÚNIOR R M C , et al. Retinal vessel segmentation using the 2-D Gabor wavelet and supervised classification. IEEE Transactions on Medical Imaging, 2006, 25 (9): 1214- 1222. doi: 10.1109/TMI.2006.879967
4	XU L , LUO S . A novel method for blood vessel detection from retinal images. Biomedical Engineering Online, 2010, 9, 14. doi: 10.1186/1475-925X-9-14
5	LUPAŞCU C A , TEGOLO D , TRUCCO E . FABC: retinal vessel segmentation using AdaBoost. IEEE Transactions on Information Technology in Biomedicine, 2010, 14 (5): 1267- 1274. doi: 10.1109/TITB.2010.2052282
6	RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C]//Proceedings of the 18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Berlin, Germany: Springer, 2015: 234-241.
7	ZHOU Z W , RAHMAN S M M , TAJBAKHSH N , et al. UNet++: a nested U-Net architecture for medical image segmentation. Berlin, Germany: Springer, 2018: 3- 11.
8	OKTAY O, SCHLEMPER J, LE FOLGOC L, et al. Attention U-Net: learning where to look for the pancreas[EB/OL]. [2024-07-11]. https://arxiv.org/abs/1804.03999v3.
9	IBTEHAZ N , RAHMAN M S . MultiResUNet: rethinking the U-Net architecture for multimodal biomedical image segmentation. Neural Networks, 2020, 121, 74- 87. doi: 10.1016/j.neunet.2019.08.025
10	GU R , WANG G , SONG T , et al. CA-Net: comprehensive attention convolutional neural networks for explainable medical image segmentation. IEEE Transactions on Medical Imaging, 2021, 40 (2): 699- 711. doi: 10.1109/TMI.2020.3035253
11	WANG Z Y, ZHENG J Q, ZHANG Y C, et al. Mamba-UNet: UNet-like pure visual mamba for medical image segmentation[EB/OL]. [2024-07-11]. https://arxiv.org/abs/2402.05079v2.
12	SINGADKAR G , MAHAJAN A , THAKUR M , et al. Deep deconvolutional residual network based automatic lung nodule segmentation. Journal of Digital Imaging, 2020, 33 (3): 678- 684. doi: 10.1007/s10278-019-00301-4
13	NI J J , WU J H , WANG H Y , et al. Global channel attention networks for intracranial vessel segmentation. Computers in Biology and Medicine, 2020, 118, 103639. doi: 10.1016/j.compbiomed.2020.103639
14	WANG J , SUN K , CHENG T , et al. Deep high-resolution representation learning for visual recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43 (10): 3349- 3364. doi: 10.1109/TPAMI.2020.2983686
15	杜妮妮, 单凯东, 王建超. HRformer: 基于多级回归Transformer网络的红外小目标检测. 红外技术, 2024, 46 (2): 199- 207.
	DU N N , SHAN K D , WANG J C . HRformer: hierarchical regression Transformer for infrared small-target detection. Infrared Technology, 2024, 46 (2): 199- 207.
16	圣文顺, 余熊峰, 林佳燕, 等. 融合注意力与特征金字塔的小尺度目标检测算法. 计算机工程, 2024, 50 (1): 242- 250. doi: 10.19678/j.issn.1000-3428.0066724
	SHENG W S , YU X F , LIN J Y , et al. Small-scale object detection algorithm integrating attention and feature Pyramids. Computer Engineering, 2024, 50 (1): 242- 250. doi: 10.19678/j.issn.1000-3428.0066724
17	CHEN L C, ZHU Y K, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of the European Conference on Computer Vision. Berlin, Germany: Springer, 2018: 833-851.
18	DAI J F, QI H Z, XIONG Y W, et al. Deformable convolutional networks[C]//Proceedings of the IEEE International Conference on Computer Vision (ICCV). Washington D.C., USA: IEEE Press, 2017: 764-773.
19	XIONG Y W, LI Z Q, CHEN Y T, et al. Efficient deformable ConvNets: rethinking dynamic and sparse operator for vision applications[EB/OL]. [2024-07-11]. https://arxiv.org/abs/2401.06197v1.
20	徐晓峰, 黄韫栀, 徐军. 基于各向异性注意力的双分支血管分割模型. 计算机工程, 2024, 50 (1): 348- 356. doi: 10.19678/j.issn.1000-3428.0067078
	XU X F , HUANG Y Z , XU J . Dual-branch vascular segmentation model based on anisotropic attention. Computer Engineering, 2024, 50 (1): 348- 356. doi: 10.19678/j.issn.1000-3428.0067078
21	DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. An image is worth 16×16 words: Transformers for image recognition at scale[EB/OL]. [2024-07-11]. https://arxiv.org/abs/2010.11929.
22	VASWANI A , SHAZEER N , PARMAR N , et al. Attention is all you need. Advances in Neural Information Processing Systems, 2017, 30, 5998- 6008.
23	CHEN J N, LU Y Y, YU Q H, et al. TransUNet: transformers make strong encoders for medical image segmentation[EB/OL]. [2024-07-11]. https://arxiv.org/abs/2102.04306v1.
24	LI J Y , GAO G , YANG L , et al. DPF-Net: a dual-path progressive fusion network for retinal vessel segmentation. IEEE Transactions on Instrumentation and Measurement, 2023, 72, 1- 17.
25	ZHANG M , YU F , ZHAO J , et al. BEFD: boundary enhancement and feature denoising for vessel segmentation. Berlin, Germany: Springer, 2020.
26	GAO Z , ZONG Q , WANG Y , et al. Laplacian salience-gated feature pyramid network for accurate liver vessel segmentation. IEEE Transactions on Medical Imaging, 2023, 42 (10): 3059- 3068. doi: 10.1109/TMI.2023.3273528
27	IOFFE S, SZEGEDY C, IOFFE S, et al. Batch normalization[C]//Proceedings of the 32nd International Conference on Machine Learning. New York, USA: ACM Press, 2015: 448-456.
28	XU S B, ZHENG S C, XU W H, et al. HCF-Net: hierarchical context fusion network for infrared small object detection[EB/OL]. [2024-07-11]. https://arxiv.org/abs/2403.10778v1.
29	SHI B F, GAI S Y, DARRELL T, et al. TOAST: transfer learning via attention steering[EB/OL]. [2024-07-11]. https://arxiv.org/abs/2305.15542v2.
30	WANG Q L, WU B G, ZHU P F, et al. ECA-Net: efficient channel attention for deep convolutional neural networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Washington D.C., USA: IEEE Press, 2020: 11534-11542.
31	SOLER L, HOSTETTLER A, AGNUS V, et al. 3D image reconstruction for comparison of algorithm database: a patient specific anatomical and medical image database[EB/OL]. [2024-07-11]. https://www.ircad.fr/research/data-sets/liver-segmentation-3d-ircadb-01/.
32	CHEN L C , PAPANDREOU G , KOKKINOS I , et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40 (4): 834- 848. doi: 10.1109/TPAMI.2017.2699184

[1]	赵南南, 高翡晨. 基于改进YOLOv8的交通场景实例分割算法[J]. 计算机工程, 2025, 51(1): 198-207.
[2]	李刚, 邵瑞, 周鸣乐, 李敏, 万洪林. 基于注意力的轻量级工业产品缺陷检测网络[J]. 计算机工程, 2023, 49(11): 275-283.
[3]	秦鹏, 唐川明, 刘云峰, 张建林, 徐智勇. 基于改进YOLOv3的红外目标检测方法[J]. 计算机工程, 2022, 48(3): 211-219.
[4]	黄帅, 张毅. 基于梯形跨尺度特征耦合网络的SAR图像舰船检测[J]. 计算机工程, 2022, 48(12): 270-280.
[5]	刘振国, 李钊, 宋滕滕, 何益智. 结合可变形卷积与双边网格的立体匹配网络[J]. 计算机工程, 2022, 48(12): 241-247,254.
[6]	尚佳童, 雷涛, 张栋, 杜晓刚, 翟钰杰. 面向刻蚀图像分割的轻量可变形编解码网络[J]. 计算机工程, 2022, 48(12): 203-211,217.
[7]	李富豪, 赵希梅. 基于D-Unet神经网络的鼻腔鼻窦肿瘤分割算法[J]. 计算机工程, 2022, 48(1): 281-287.
[8]	朱灵灵, 高超, 陈福才. 基于轻量级卷积神经网络的人脸检测算法[J]. 计算机工程, 2021, 47(7): 273-280.
[9]	包俊, 刘宏哲. 融合可变形卷积网络的鱼眼图像目标检测[J]. 计算机工程, 2021, 47(4): 248-255.
[10]	喻清挺, 喻维超, 喻国平. 基于改进R-FCN的交通标志检测[J]. 计算机工程, 2021, 47(12): 285-290,298.
[11]	黄凤琪, 陈明, 冯国富. 基于可变形卷积的改进YOLO目标检测算法[J]. 计算机工程, 2021, 47(10): 269-275,282.
[12]	翟强, 王陆洋, 殷保群, 彭思凡, 邢思思. 基于尺度自适应卷积神经网络的人群计数算法[J]. 计算机工程, 2020, 46(2): 250-254,261.

选择文件类型/文献管理软件名称

选择包含的内容