EchoGPK: A Lightweight Graph Convolutional Analysis Method for Echocardiography Based on Prior Knowledge Guidance

WANG Bo-rong; YE Jian

doi:10.12263/DZXB.20231065

您当前的位置：

首页 >

文章列表页 >

EchoGPK: A Lightweight Graph Convolutional Analysis Method for Echocardiography Based on Prior Knowledge Guidance

PAPERS | 更新时间：2025-12-08

- EchoGPK: A Lightweight Graph Convolutional Analysis Method for Echocardiography Based on Prior Knowledge Guidance
- ACTA ELECTRONICA SINICA Vol. 52, Issue 4, Pages: 1296-1304(2024)
- 作者机构：
  
  1.中国科学院计算技术研究所,北京 100190
  2.中国科学院大学,北京 101408
  3.移动计算与新型终端北京市重点实验室,北京 100190
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(82374299);National Key Research and Development Program of China(2022YFB3904700)
- DOI：10.12263/DZXB.20231065
  CLC： TP391.41
- Received：15 November 2023，
  
  Revised：2024-02-04，
  
  Published：25 April 2024
- 稿件说明：
移动端阅览
王博荣, 叶剑. EchoGPK:基于先验知识引导的超声心动图轻量级图卷积分析方法[J]. 电子学报, 2024, 52(04): 1296-1304.

WANG Bo-rong, YE Jian. EchoGPK: A Lightweight Graph Convolutional Analysis Method for Echocardiography Based on Prior Knowledge Guidance[J]. Acta Electronica Sinica, 2024, 52(04): 1296-1304.
王博荣, 叶剑. EchoGPK:基于先验知识引导的超声心动图轻量级图卷积分析方法[J]. 电子学报, 2024, 52(04): 1296-1304. DOI：10.12263/DZXB.20231065

WANG Bo-rong, YE Jian. EchoGPK: A Lightweight Graph Convolutional Analysis Method for Echocardiography Based on Prior Knowledge Guidance[J]. Acta Electronica Sinica, 2024, 52(04): 1296-1304. DOI：10.12263/DZXB.20231065

摘要

根据超声心动图准确分析左心室轮廓和射血分数对于心血管疾病诊断意义重大.但现有方法存在左心室分割和射血分数预测之间缺乏关联性、左心室分割关键点易于出现离群点和突变点、方法存储和计算开销大、解释性不佳等问题，为此提出一种基于先验知识引导的轻量级图卷积方法EchoGPK（Echo Guided by Priori Knowledge），以心脏的结构和运动特性、相邻心肌的相似性等先验知识为引导，设计了计算高效的螺旋聚合函数和深度压缩的多头偏心聚合解码器，实现了图卷积结构的轻量化.方法基于临床医生的普遍经验提出了适度利用左心室轮廓的多任务射血分数预测网络，建立了左心室分割和射血分数预测之间的关联性，增强了推理的可解释性；基于图卷积神经网络的传递特性约束邻居点的行为，减少了边界离群点和突变点的产生.EchoGPK在大型公开数据集EchoNet-Dynamic上的实验结果表明，左心室分割的Dice分数达92.13%，射血分数预测的MAE达3.92%；方法表现出准确率高、参数量和算力需求低等特点，证明了先验知识在超声医学图像分析中的有效性.

Abstract

Accurate analysis of the left ventricular outline and ejection fraction through echocardiography holds significant diagnostic implications in cardiovascular diseases. However

current methodologies exhibit deficiencies such as a lack of correlation between left ventricular segmentation and ejection fraction prediction

susceptibility to outliers and abrupt variations in key points of left ventricular segmentation

substantial storage and computational overhead

and poor interpretability. In addressing these issues

this study proposes a lightweight graph convolutional method termed EchoGPK (Echo Guided by Priori Knowledge). Guided by prior knowledge encompassing cardiac structure

motion characteristics

and the similarity among adjacent myocardial regions

the approach incorporates a computationally efficient spiral aggregation function and a deeply compressed multi-head eccentric aggregation decoder

achieving the lightweighting of the graph convolutional structure. Leveraging the common experiences of clinical practitioners

the method introduces a multi task ejection fraction prediction network that moderately utilizes left ventricular contours

establishing a correlation between left ventricular segmentation and ejection fraction prediction to enhance interpretability. By employing the graph convolutional neural network transmission characteristics to constrain the behavior of neighboring points

the generation of boundary outliers and abrupt variations is reduced. Experimental results on the large-scale public dataset EchoNet-Dynamic demonstrate that EchoGPK achieves a Dice score of 92.13% for left ventricular segmentation and a mean absolute error (MAE) of 3.92% for ejection fraction prediction. Furthermore

the method exhibits higher accuracy

superior parameter count and computational efficiency compared to relevant approaches

affirming the effectiveness of prior knowledge in ultrasound medical image analysis.

关键词

Keywords

references

THOMAS S , GILBERT A , BEN-YOSEF G . Light-weight spatio-temporal graphs for segmentation and ejection fraction prediction in cardiac ultrasound [C ] // Medical Image Computing and Computer Assisted Intervention—MICCAI 2022: 25th International Conference . Cham : Springer Nature Switzerland , 2022 : 380 - 390 .

刘盛锋 , 王毅 , 杨鑫 , 等 . 深度学习在医学超声图像分析中的应用综述 [J ] . 工程(英文) , 2019 , 5 ( 2 ): 261 - 275 .

LIU S F , WANG Y , YANG X , et al . Deep learning in medical ultrasound analysis: A review [J ] . Engineering , 2019 , 5 ( 2 ): 261 - 275 . (in Chinese)

RONNEBERGER O , FISCHER P , BROX T . U-net: Convolutional networks for biomedical image segmentation [C ] // Medical Image Computing and Computer-Assisted Intervention—MICCAI 2015: 18th International Conference . Cham : Springer International Publishing , 2015 : 234 - 241 .

ISENSEE F , JAEGER P F , KOHL S A A , et al . nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation [J ] . Nature Methods , 2021 , 18 ( 2 ): 203 - 211 .

OUYANG D , HE B , GHORBANI A , et al . Video-based AI for beat-to-beat assessment of cardiac function [J ] . Nature , 2020 , 580 ( 7802 ), 252 - 256 .

MUHTASEB R , YAQUB M . EchoCoTr: Estimation of the left ventricular ejection fraction from spatiotemporal echocardiography [C ] // Medical Image Computing and Computer Assisted Intervention—MICCAI 2022: 25th International Conference . Cham : Springer Nature Switzerland , 2022 : 370 - 379 .

ZHANG Y D , LIU H Y , HU Q . TransFuse: Fusing Transformers and CNNs for medical image segmentation [C ] // Medical Image Computing and Computer Assisted Intervention—MICCAI 2021: 24th International Conference . Cham : Springer Nature Switzerland , 2021 : 14 - 24 .

XUE W F , CAO H , MA J Q , et al . Improved segmentation of echocardiography with orientation-congruency of optical flow and motion-enhanced segmentation [J ] . IEEE Journal of Biomedical and Health Informatics , 2022 , 26 ( 12 ): 6105 - 6115 .

刘晓鸣 , 雷震 , 何刊 , 等 . 全卷积神经网络与全连接条件随机场中的左心室射血分数精准计算 [J ] . 计算机辅助设计与图形学学报 , 2019 , 31 ( 3 ): 431 - 438 .

LIU X M , LEI Z , HE K , et al . Accurate estimation of left ventricle ejection fraction using fully convolutional networks and fully connected conditional random field [J ] . Journal of Computer-Aided Design & Computer Graphics , 2019 , 31 ( 3 ): 431 - 438 . (in Chinese)

BOURITSAS G , BOKHNYAK S , PLOUMPIS S , et al . Neural 3D morphable models: Spiral convolutional networks for 3D shape representation learning and generation [C ] // 2019 IEEE/CVF International Conference on Computer Vision (ICCV) . Piscataway : IEEE , 2019 : 7212 - 7221 .

万升 , 杨健 , 宫辰 . 基于图神经网络的高光谱图像分类研究进展 [J ] . 电子学报 , 2023 , 51 ( 6 ): 1687 - 1709 .

WAN S , YANG J , GONG C . Advances of hyperspectral image classification based on graph neural networks [J ] . Acta Electronica Sinica , 2023 , 51 ( 6 ): 1687 - 1709 . (in Chinese)

SANDLER M , HOWARD A , ZHU M L , et al . MobileNetV2: Inverted residuals and linear bottlenecks [C ] // 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2018 : 4510 - 4520 .

HE K M , ZHANG X Y , REN S Q , et al . Deep residual learning for image recognition [C ] // 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . Piscataway : IEEE , 2016 : 770 - 778 .

TRAN D , WANG H , TORRESANI L , et al . A closer look at spatiotemporal convolutions for action recognition [C ] // 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2018 : 6450 - 6459 .

ESFEH M M K , LUONG C , BEHNAMI D , et al . A deep bayesian video analysis framework: Towards a more robust estimation of ejection fraction [C ] //Medical Image Computing and Computer Assisted Intervention—MICCAI 2020: The 23rd International Conference , Part II . New York : ACM , 2020 : 582 - 590 .

REYNAUD H , VLONTZOS A , HOU B , et al . Ultrasound video Transformers for cardiac ejection fraction estimation [C ] // Medical Image Computing and Computer Assisted Intervention—MICCAI 2021: The 24th International Conference . Cham : Springer Nature Switzerland , 2021 : 495 - 505 .

ZHOU Z W , RAHMAN SIDDIQUEE M M , TAJBAKHSH N . UNet++: A nested U-Net architecture for medical image segmentation [C ] // Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support (DLMIA ML-CDS) 2018 . Cham : Springer Nature Switzerland , 2018 : 3 - 11 .

TIAN Z Q , LI X J , ZHENG Y Y , et al . Graph-convolutional-network-based interactive prostate segmentation in MR images [J ] . Medical Physics , 2020 , 47 ( 9 ): 4164 - 4176 .

DAI W H , LI X M , CHIU W H K , et al . Adaptive contrast for image regression in computer-aided disease assessment [J ] . IEEE Transactions on Medical Imaging , 2022 , 41 ( 5 ): 1255 - 1268 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Graph-Based Representation Knowledge Distillation for Image Classification

A Deep Semantic Mining Model Based on Aspect-Level Sentiment Analysis

Indoor Through-The-Wall Passive Target Detection Algorithm Based on Graph Convolutional Neural Network

Related Author

XU Yong-jun

YANG Chuan-guang

CHEN Lu-ming

ZHAO Er-hu

AN Zhu-lin

PENG Jun-jie

ZHANG Huan-xiang

YANG Xiao-long

Related Institution

Unit of PLA

School of Computer Engineering and Science，Shanghai University

Shanghai Institute for Advanced Communication and Data Science

School of Innovation and Entrepreneurship Education， Inner Mongolia University of Science and Technology

School of Communication and Information Engineering， Chongqing， University of Posts and Telecommunications

⁰