Centerline Tracking Algorithm for Aortic Dissection Based on Deep Reinforcement Learning

doi:10.19678/j.issn.1000-3428.0070257

Computer Engineering ›› 2026, Vol. 52 ›› Issue (6): 414-424. doi: 10.19678/j.issn.1000-3428.0070257

• Interdisciplinary Integration and Engineering Applications • Previous Articles

Centerline Tracking Algorithm for Aortic Dissection Based on Deep Reinforcement Learning

ZENG An¹, ZHENG Jiayu¹, PAN Dan², ZHAO Jingliang¹^,*(), HUANG Xingqing³

1. School of Computer Science, Guangdong University of Technology, Guangzhou 510006, Guangdong, China
2. School of Electronics and Information, Guangdong Polytechnic Normal University, Guangzhou 510665, Guangdong, China
3. The Affiliated Panyu Central Hospital, Guangzhou Medical University, Guangzhou 511431, Guangdong, China

Received:2024-08-15 Revised:2024-10-23 Online:2026-06-15 Published:2025-01-03
Contact: ZHAO Jingliang

基于深度强化学习的主动脉夹层中心线追踪算法

曾安¹, 郑嘉裕¹, 潘丹², 赵靖亮¹^,*(), 黄幸青³

1. 广东工业大学计算机学院, 广东广州 510006
2. 广东技术师范大学电子与信息学院, 广东广州 510665
3. 广州医科大学附属番禺中心医院, 广东广州 511431

通讯作者: 赵靖亮
作者简介:
曾安(CCF杰出会员), 教授、博士, 主研方向为图像处理、数据挖掘
郑嘉裕, 硕士研究生
潘丹, 教授、博士
赵靖亮(通信作者), 讲师、博士
黄幸青, 主任医师
基金资助:
广东省科技计划项目(2019A050510041); 广东省自然科学基金(2021A1515012300); 国家自然科学基金(61976058); 国家自然科学基金(92267107)

Abstract

Abstract:

The extraction of the Aortic Dissection (AD) centerline is crucial for its quantitative diagnosis and treatment. However, owing to the complex anatomy and diverse vascular morphology of AD, this task is highly challenging, and current quantitative evaluation methods are limited. Most existing approaches require presegmentation or distance map computation, followed by extraction using minimum path or skeleton algorithms; however, these often result in broken centerlines owing to incomplete segmentation. We propose a Deep Q-Network (DQN)-based centerline tracking algorithm that integrates an attention-embedded dilated residual module with a channel attention mechanism, thereby enabling more effective vascular feature extraction and automatic tracking of complex vessel centerlines. Additionally, an improved reward function is designed to guide accurate centerline tracking. Experiments on public datasets show that our method outperforms others in terms of centerline overlap metrics, with an average extraction speed of 5 s per case, indicating its strong clinical potential.

Key words: Aortic Dissection (AD), centerline extraction, Deep Q-Network (DQN), Deep Reinforcement Learning (DRL), attention mechanism

摘要：

主动脉夹层(AD)中心线的提取在AD疾病的定量诊断和治疗中具有极其重要的临床意义。然而, 由于AD的解剖结构复杂及血管形态和病变区域多样化等因素, AD中心线的提取任务非常具有挑战性, 且目前对这一任务的定量评估研究仍然较为有限。当前多数方法在提取中心线时, 需要进行预先分割、全卷扫描操作或计算距离图, 随后使用最小路径或骨架算法进行提取。然而, AD腔体难以完整分割, 上述方法所得中心线易存在断裂。为此, 提出一种基于深度Q网络(DQN)的中心线跟踪算法, 并设计一个注意力嵌入的空洞残差模块, 将其与通道注意力机制结合, 能够更有效地提取血管特征并自动追踪复杂病变血管的中心线。此外, 提出一种改进的奖励函数, 引导智能体准确地追踪中心线。在公开数据集上的实验结果表明, 提出的方法在中心线重叠度指标上全面优于对比算法, 提取一例数据中心线的平均速度为5 s, 具有较好的临床应用潜力。

关键词: 主动脉夹层, 中心线提取, 深度Q网络, 深度强化学习, 注意力机制

ZENG An, ZHENG Jiayu, PAN Dan, ZHAO Jingliang, HUANG Xingqing. Centerline Tracking Algorithm for Aortic Dissection Based on Deep Reinforcement Learning[J]. Computer Engineering, 2026, 52(6): 414-424.

曾安, 郑嘉裕, 潘丹, 赵靖亮, 黄幸青. 基于深度强化学习的主动脉夹层中心线追踪算法[J]. 计算机工程, 2026, 52(6): 414-424.

/ Recommend / Download Citations

URL: https://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0070257

https://www.ecice06.com/EN/Y2026/V52/I6/414

Figures/Tables 12

Fig.1 Procedure of centerline tracking algorithm

Fig.2 Network structure for centerline tracking of aortic dissection

Fig.3 The intuitive perspective of the reward function

Fig.4 Centerline tracking results of aortic dissection

Fig.5 Boxplots of the evaluation metrics on the training set and the test set

Fig.6 Examples of centerlines extracted of different methods

Fig.7 Comparison of reward function optimization before and after in the proposed method

References 29

1	ELMARASI M , ELMAKATY I , ELSAYED B , et al. Phenotypic switching of vascular smooth muscle cells in atherosclerosis, hypertension, and aortic dissection. Journal of Cellular Physiology, 2024, 38, 239- 247.
2	BALIYAN V , PARAKH A , PRABHAKAR A M , et al. Acute aortic syndromes and aortic emergencies. Cardiovascular Diagnosis & Therapy, 2018, 8 (S1): 82- 96.
3	PAPE L A , AWAIS M , WOZNICKI E M . Presentation, diagnosis, and outcomes of acute aortic dissection: seventeen-year trends from the international registry of acute aortic dissection. Journal of Vascular Surgery, 2016, 23 (2): 552- 553.
4	CAMBRIA R P . Surgical treatment of complicated distal aortic dissection. Seminars in Vascular Surgery, 2002, 15 (2): 97- 107. doi: 10.1053/svas.2002.33439
5	YIN J , LIU F , WANG J , et al. Aortic dissection: global epidemiology. Cardiology Plus, 2022, 7 (4): 151- 161.
6	张涛. 主动脉夹层64例临床分析. 哈尔滨医药, 2024, 44 (2): 57- 59.
	ZHANG T . Clinical analysis of 64 cases of aortic dissection. Harbin Medicine, 2024, 44 (2): 57- 59.
7	BOIX-GARIBO R , UZZAMAN M M , BAPAT V N . Review of minimally invasive aortic valve surgery. Interventional Cardiology, 2015, 10 (3): 144- 148.
8	ZHANG P, WANG F, ZHENG Y. Deep reinforcement learning for vessel centerline tracing in multi-modality 3D volumes[C]//Proceedings of the International Conference on Medical Image Computing and Computer-assisted intervention. Berlin, Germany: Springer, 2018: 223-231.
9	LI Z, XIA Q, HU Z, et al. A deep reinforced tree-traversal agent for coronary artery centerline extraction[C]//Proceedings of the 24th International Conference on Medical Image Computing and Computer Assisted Intervention. Berlin, Germany: Springer, 2021: 418-428.
10	BIAN Z , TAN W , YANG J , et al. Accurate airway centerline extraction based on topological thinning using graph-theoretic analysis. Bio-medical Materials and Engineering, 2014, 24 (6): 3239- 3249. doi: 10.3233/BME-141146
11	ZHOU C , CHAN H P , CHUGHTAI A , et al. Automated coronary artery tree extraction in coronary CT angiography using a multiscale enhancement and dynamic balloon tracking method. Computerized Medical Imaging and Graphics, 2012, 36 (1): 1- 10. doi: 10.1016/j.compmedimag.2011.04.001
12	WOLTERINK J M , VAN HAMERSVELT R W , VIERGEVER M A , et al. Coronary artery centerline extraction in cardiac CT angiography using a CNN-based orientation classifier. Medical Image Analysis, 2019, 51, 46- 60. doi: 10.1016/j.media.2018.10.005
13	郑会吉, 余思聪, 邱鑫源, 等. 基于元强化学习的自适应卸载方法. 电讯技术, 2024, 64 (2): 177- 183.
	ZHENG H J , YU S C , QIU X Y , et al. An adaptive offloading method based on meta reinforcement learning. Telecommunications Engineering, 2024, 64 (2): 177- 183.
14	刘全, 翟建伟, 章宗长, 等. 深度强化学习综述. 计算机学报, 2018, 41 (1): 1- 27.
	LIU Q , ZHAI J W , ZHANG Z Z , et al. A survey on deep reinforcement learning. Chinese Journal of Computers, 2018, 41 (1): 1- 27.
15	ZHANG Y, LUO G, WANG W, et al. Branch-aware double DQN for centerline extraction in coronary CT angiography[C]//Proceedings of the 23rd International Conference on Medical Image Computing and Computer Assisted Intervention. Berlin, Germany: Springer, 2020: 35-44.
16	HASSELT H V, GUEZ A, SILVER D. Deep reinforcement learning with double Q-learning[EB/OL]. [2024-07-10]. https://arxiv.org/abs/1509.06461.
17	MNIH V , KAVUKCUOGLU K , SILVER D , et al. Human-level control through deep reinforcement learning. Nature, 2015, 518 (7540): 529- 533. doi: 10.1038/nature14236
18	SU J , LI S , WOLFF L , et al. Deep reinforcement learning for cerebral anterior vessel tree extraction from 3D CTA images. Medical Image Analysis, 2023, 84, 102724. doi: 10.1016/j.media.2022.102724
19	SCHULMN J, WOLSKI F, DHARIEAL P, et al. Proximal policy optimization algorithms[EB/OL]. [2024-07-10]. https://arxiv.org/abs/1707.06347.
20	DAYAN P , WATKINS C J C H . Reinforcement learning. Machine Learning, 1992, 8 (3): 279- 292.
21	MNIH V , KAVUKCUOGLU K , SILVER D , et al. Playing atari with deep reinforcement learning. Computer Science, 2013, 45, 457- 466.
22	WANG Z, SCHAUL T, HESSEL M, et al. Dueling network architectures for deep reinforcement learning[C]//Proceedings of the International Conference on Machine Learning. Washington D. C., USA: IEEE Press, 2016: 1995-2003.
23	YU F , KOLTUN V , FUNKHOUSER T . Dilated residual networks. IEEE Computer Society, 2017, 42 (7): 563- 572.
24	HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Washington D. C., USA: IEEE Press, 2018: 7132-7141.
25	YAO Z , XIE W , ZHANG J , et al. Imagetbad: a 3d computed tomography angiography image dataset for automatic segmentation of type-b aortic dissection. Frontiers in Physiology, 2021, 12, 732711. doi: 10.3389/fphys.2021.732711
26	ZHAO J , FENG Q . Automatic aortic dissection centerline extraction via morphology-guided CRN tracker. IEEE Journal of Biomedical and Health Informatics, 2023, 25 (9): 3473- 3485.
27	BALOCCO S , GATTA C , CIOMPI F , et al. Standardized evaluation methodology and reference database for evaluating IVUS image segmentation. Computerized Medical Imaging and Graphics, 2023, 38 (2): 70- 90.
28	VALANARASU J M J, PATEL V M. UNeXt: MLP-based rapid medical image segmentation network[C]//Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention. Berlin, Germany: Springer, 2022: 23-33.
29	GRELARD F , BALDACCI F , VIALARD A , et al. New methods for the geometrical analysis of tubular organs. Medical Image Analysis, 2023, 42 (5): 89- 101.

[1]	ZHOU Lijun, ZHANG Junran, WANG Kaiyuan, XIANG Junlian. Drug Recommendation Based on Graph-Enhanced Attention Integrating Clinical Features of Patients [J]. Computer Engineering, 2026, 52(6): 314-325.
[2]	DAI Yinqiao, XIAO Wulong, LI Bailin, LI Li. Lettuce Core Detection Algorithm Based on Improved YOLOv5s [J]. Computer Engineering, 2026, 52(6): 352-364.
[3]	HU Kangyuan, GUO Tao, MU Nan. Image Inpainting Method for Cultural Relics Based on Self-Attention Mechanism and Dynamic Masking Mechanism [J]. Computer Engineering, 2026, 52(6): 179-188.
[4]	WANG Yongqi, WANG Lei. Multimodal Sentiment Recognition Based on Cross-Modal Enhancement and Time-Step Gating [J]. Computer Engineering, 2026, 52(6): 258-267.
[5]	YU Mengyuan, LIU Xiangyang. Ship Detection Method Based on Multimodal Visible and Infrared Image Fusion [J]. Computer Engineering, 2026, 52(6): 278-287.
[6]	LUO Heng, WAN Liang. Network Traffic Intrusion Detection Method Based on Dynamic Spatio-Temporal Graph Neural Network [J]. Computer Engineering, 2026, 52(6): 202-213.
[7]	XIAO Zeqiu, LI Yong, WANG Xia. Prediction of Blood Glucose Concentration in Diabetic Patients Based on PBI-CLA Model [J]. Computer Engineering, 2026, 52(6): 382-390.
[8]	ZHAO Shuxu, ZHOU Hongze, WANG Xiaolong. Optimization of Optimal Coalition Structure Generation Strategy Based on Improved DQN [J]. Computer Engineering, 2026, 52(5): 117-128.
[9]	WU Peiying, LI Xiaohui, WANG Junfeng. C2 Traffic Detection Based on Context-aware Language Model [J]. Computer Engineering, 2026, 52(5): 270-280.
[10]	QU Jinghong, WANG Zhongqing, ZHOU Guodong. Knowledge Text Generation Based on Pre-trained Model for Question and Answer [J]. Computer Engineering, 2026, 52(5): 326-335.
[11]	LI Mingming, PAN Zihao. Mobile Robot Path Planning Based on Improved TD3 Algorithm [J]. Computer Engineering, 2026, 52(5): 150-159.
[12]	SONG Tianze, CAO Congjun, HE Jiaqi, WANG Xusheng, LIU Chenyu. Research on Dense Pedestrian Detection Algorithm Based on Improved DETR [J]. Computer Engineering, 2026, 52(5): 250-258.
[13]	YANG Jiahao, WANG Lei. Individual Attention Target Detection Based on Multi-Feature Spatiotemporal Inference Network [J]. Computer Engineering, 2026, 52(5): 184-191.
[14]	ZHANG Hong, ZHU Siyu, ZHANG Xijun, WEI Jiaoyun. Research on Non-Stationary Traffic Flow Forecasting Based on Adaptive Graph Convolution with Optimized Meta-Graph Learning [J]. Computer Engineering, 2026, 52(5): 456-466.
[15]	LI Jiao, FAN Haodong, HONG Xudong, XU Zhenyi, FAN Xu, HUANG Jun. Multi-Label Image Classification Based on Label Visual Prototype Learning [J]. Computer Engineering, 2026, 52(4): 229-238.

Please choose a citation manager

Content to export