• 文献检索
  • 文档翻译
  • 深度研究
  • 学术资讯
  • Suppr Zotero 插件Zotero 插件
  • 邀请有礼
  • 套餐&价格
  • 历史记录
应用&插件
Suppr Zotero 插件Zotero 插件浏览器插件Mac 客户端Windows 客户端微信小程序
定价
高级版会员购买积分包购买API积分包
服务
文献检索文档翻译深度研究API 文档MCP 服务
关于我们
关于 Suppr公司介绍联系我们用户协议隐私条款
关注我们

Suppr 超能文献

核心技术专利：CN118964589B侵权必究
粤ICP备2023148730 号-1Suppr @ 2026

文献检索

告别复杂PubMed语法，用中文像聊天一样搜索，搜遍4000万医学文献。AI智能推荐，让科研检索更轻松。

立即免费搜索

文件翻译

保留排版，准确专业，支持PDF/Word/PPT等文件格式，支持 12+语言互译。

免费翻译文档

深度研究

AI帮你快速写综述，25分钟生成高质量综述，智能提取关键信息，辅助科研写作。

立即免费体验

CDKD-w+：一种基于w+空间编码的冠状动脉数字减影血管造影视频序列关键帧识别方法。

CDKD-w+: A Keyframe Recognition Method for Coronary Digital Subtraction Angiography Video Sequence Based on w+ Space Encoding.

作者信息

Zhu Yong, Li Haoyu, Xiao Shuai, Yu Wei, Shang Hongyu, Wang Lin, Liu Yang, Wang Yin, Yang Jiachen

机构信息

School of Electrical and Information Engineering, Tianjin University, Tianjin 300072, China.

Tianjin Institute of Software Engineering, Tianjin 300387, China.

出版信息

Sensors (Basel). 2025 Jan 24;25(3):710. doi: 10.3390/s25030710.

DOI:10.3390/s25030710
PMID:39943348
原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC11821101/
Abstract

Currently, various deep learning methods can assist in medical diagnosis. Coronary Digital Subtraction Angiography (DSA) is a medical imaging technology used in cardiac interventional procedures. By employing X-ray sensors to visualize the coronary arteries, it generates two-dimensional images from any angle. However, due to the complexity of the coronary structures, the 2D images may sometimes lack sufficient information, necessitating the construction of a 3D model. Camera-level 3D modeling can be realized based on deep learning. Nevertheless, the beating of the heart results in varying degrees of arterial vasoconstriction and vasodilation, leading to substantial discrepancies between DSA sequences, which introduce errors in 3D modeling of the coronary arteries, resulting in the inability of the 3D model to reflect the coronary arteries. We propose a coronary DSA video sequence keyframe recognition method, CDKD-w+, based on w+ space encoding. The method utilizes a pSp encoder to encode the coronary DSA images, converting them into latent codes in the w+ space. Differential analysis of inter-frame latent codes is employed for heartbeat keyframe localization, aiding in coronary 3D modeling. Experimental results on a self-constructed coronary DSA heartbeat keyframe recognition dataset demonstrate an accuracy of 97%, outperforming traditional metrics such as L1, SSIM, and PSNR.

摘要

目前，各种深度学习方法可辅助医学诊断。冠状动脉数字减影血管造影（DSA）是一种用于心脏介入手术的医学成像技术。通过使用X射线传感器来可视化冠状动脉，它可以从任何角度生成二维图像。然而，由于冠状动脉结构的复杂性，二维图像有时可能缺乏足够的信息，因此需要构建三维模型。基于深度学习可以实现相机级别的三维建模。然而，心脏跳动会导致动脉不同程度的收缩和舒张，导致DSA序列之间存在显著差异，这在冠状动脉的三维建模中引入了误差，导致三维模型无法反映冠状动脉。我们提出了一种基于w+空间编码的冠状动脉DSA视频序列关键帧识别方法CDKD-w+。该方法利用pSp编码器对冠状动脉DSA图像进行编码，将其转换为w+空间中的潜在代码。通过对帧间潜在代码进行差分分析来定位心跳关键帧，辅助冠状动脉三维建模。在自建的冠状动脉DSA心跳关键帧识别数据集上的实验结果表明，准确率达到97%，优于L1、SSIM和PSNR等传统指标。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/c729b2af9861/sensors-25-00710-g014.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/e32ddeefe5cb/sensors-25-00710-g001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/76a8e4fe896c/sensors-25-00710-g002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/483e7b15339a/sensors-25-00710-g003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/b57a1a1f1dab/sensors-25-00710-g004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/66ee47c44c98/sensors-25-00710-g005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/5878432b542f/sensors-25-00710-g006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/ecebb9ea775c/sensors-25-00710-g007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/acd051710d87/sensors-25-00710-g008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/e9ffa2f6074b/sensors-25-00710-g009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/05acad0c0d1f/sensors-25-00710-g010.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/feea93dce1b1/sensors-25-00710-g011.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/e95e1e25800f/sensors-25-00710-g012.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/6f74abe70e10/sensors-25-00710-g013.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/c729b2af9861/sensors-25-00710-g014.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/e32ddeefe5cb/sensors-25-00710-g001.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/76a8e4fe896c/sensors-25-00710-g002.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/483e7b15339a/sensors-25-00710-g003.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/b57a1a1f1dab/sensors-25-00710-g004.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/66ee47c44c98/sensors-25-00710-g005.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/5878432b542f/sensors-25-00710-g006.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/ecebb9ea775c/sensors-25-00710-g007.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/acd051710d87/sensors-25-00710-g008.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/e9ffa2f6074b/sensors-25-00710-g009.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/05acad0c0d1f/sensors-25-00710-g010.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/feea93dce1b1/sensors-25-00710-g011.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/e95e1e25800f/sensors-25-00710-g012.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/6f74abe70e10/sensors-25-00710-g013.jpg
https://cdn.ncbi.nlm.nih.gov/pmc/blobs/a7f4/11821101/c729b2af9861/sensors-25-00710-g014.jpg

相似文献

1
CDKD-w+: A Keyframe Recognition Method for Coronary Digital Subtraction Angiography Video Sequence Based on w+ Space Encoding.CDKD-w+：一种基于w+空间编码的冠状动脉数字减影血管造影视频序列关键帧识别方法。
Sensors (Basel). 2025 Jan 24;25(3):710. doi: 10.3390/s25030710.
2
Training of a deep learning based digital subtraction angiography method using synthetic data.使用合成数据训练基于深度学习的数字减影血管造影方法。
Med Phys. 2024 Jul;51(7):4793-4810. doi: 10.1002/mp.16973. Epub 2024 Feb 14.
3
[Development of the DSA Method for Coronary Angiography Using Deep Learning Techniques].[基于深度学习技术的冠状动脉造影DSA方法的开发]
Nihon Hoshasen Gijutsu Gakkai Zasshi. 2022;78(2):129-139. doi: 10.6009/jjrt.780203.
4
Pretrained subtraction and segmentation model for coronary angiograms.冠状动脉造影术的预训练减影和分割模型。
Sci Rep. 2024 Aug 27;14(1):19888. doi: 10.1038/s41598-024-71063-5.
5
Subtraction-free artifact-aware digital subtraction angiography image generation for head and neck vessels from motion data.基于运动数据的无减法伪影感知数字减影血管造影术对头颈部血管的图像生成
Comput Med Imaging Graph. 2025 Apr;121:102512. doi: 10.1016/j.compmedimag.2025.102512. Epub 2025 Feb 18.
6
Deep learning-based digital subtraction angiography image generation.基于深度学习的数字减影血管造影图像生成。
Int J Comput Assist Radiol Surg. 2019 Oct;14(10):1775-1784. doi: 10.1007/s11548-019-02040-x. Epub 2019 Jul 31.
7
In vivo imaging of rat coronary arteries using bi-plane digital subtraction angiography.使用双平面数字减影血管造影术对大鼠冠状动脉进行体内成像。
J Pharmacol Toxicol Methods. 2011 Sep-Oct;64(2):151-7. doi: 10.1016/j.vascn.2011.05.008. Epub 2011 Jun 13.
8
Dual-energy CT angiography in the evaluation of intracranial aneurysms: image quality, radiation dose, and comparison with 3D rotational digital subtraction angiography.双能量 CT 血管造影在颅内动脉瘤评估中的应用：图像质量、辐射剂量，以及与三维旋转数字减影血管造影的比较。
AJR Am J Roentgenol. 2010 Jan;194(1):23-30. doi: 10.2214/AJR.08.2290.
9
[The experimental research on the frameless registration based on the digital subtraction angiography].基于数字减影血管造影的无框架配准实验研究
Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2007 Feb;24(1):23-5, 44.
10
Three-dimensional digital subtraction angiography vs two-dimensional digital subtraction angiography for detection of ruptured intracranial aneurysms: a study of 86 aneurysms.三维数字减影血管造影术与二维数字减影血管造影术用于检测破裂颅内动脉瘤的比较：86例动脉瘤的研究
Neurol India. 2005 Sep;53(3):287-9; discussion 290. doi: 10.4103/0028-3886.16925.

本文引用的文献

1
MedYOLO: A Medical Image Object Detection Framework.MedYOLO：一种医学图像目标检测框架。
J Imaging Inform Med. 2024 Dec;37(6):3208-3216. doi: 10.1007/s10278-024-01138-2. Epub 2024 Jun 6.
2
Segment anything in medical images.在医学图像中分割任何内容。
Nat Commun. 2024 Jan 22;15(1):654. doi: 10.1038/s41467-024-44824-z.
3
Generative AI-Based Data Completeness Augmentation Algorithm for Data-Driven Smart Healthcare.基于生成式人工智能的数据完整性增强算法在数据驱动型智能医疗中的应用
IEEE J Biomed Health Inform. 2025 Jun;29(6):4001-4008. doi: 10.1109/JBHI.2023.3327485.
4
An Interpretable and Accurate Deep-Learning Diagnosis Framework Modeled With Fully and Semi-Supervised Reciprocal Learning.基于全监督和半监督互惠学习的可解释且精确的深度学习诊断框架模型
IEEE Trans Med Imaging. 2024 Jan;43(1):392-404. doi: 10.1109/TMI.2023.3306781. Epub 2024 Jan 2.
5
Locating X-Ray Coronary Angiogram Keyframes via Long Short-Term Spatiotemporal Attention With Image-to-Patch Contrastive Learning.通过具有图像到补丁对比学习的长短期时空注意力定位X射线冠状动脉造影关键帧
IEEE Trans Med Imaging. 2024 Jan;43(1):51-63. doi: 10.1109/TMI.2023.3286859. Epub 2024 Jan 2.
6
Federated Active Learning for Multicenter Collaborative Disease Diagnosis.联邦主动学习在多中心协作疾病诊断中的应用。
IEEE Trans Med Imaging. 2023 Jul;42(7):2068-2080. doi: 10.1109/TMI.2022.3227563. Epub 2023 Jun 30.
7
Progressive Perception Learning for Main Coronary Segmentation in X-Ray Angiography.用于X射线血管造影中主要冠状动脉分割的渐进式感知学习
IEEE Trans Med Imaging. 2023 Mar;42(3):864-879. doi: 10.1109/TMI.2022.3219126. Epub 2023 Mar 2.
8
Self-supervised learning enables 3D digital subtraction angiography reconstruction from ultra-sparse 2D projection views: A multicenter study.基于自监督学习的超稀疏二维投影视图三维数字减影血管造影重建：一项多中心研究。
Cell Rep Med. 2022 Oct 18;3(10):100775. doi: 10.1016/j.xcrm.2022.100775. Epub 2022 Oct 7.
9
A digital mask to safeguard patient privacy.数字口罩保护患者隐私。
Nat Med. 2022 Sep;28(9):1883-1892. doi: 10.1038/s41591-022-01966-1. Epub 2022 Sep 15.
10
CAR-Net: A Deep Learning-Based Deformation Model for 3D/2D Coronary Artery Registration.CAR-Net：一种基于深度学习的三维/二维冠状动脉配准变形模型。
IEEE Trans Med Imaging. 2022 Oct;41(10):2715-2727. doi: 10.1109/TMI.2022.3168786. Epub 2022 Sep 30.