[1]张建华,刘新科,赵岩,等.基于3D-RA图像的颅内动脉瘤自动检测算法[J].中国医学物理学杂志,2022,39(8):950-956.[doi:DOI:10.3969/j.issn.1005-202X.2022.08.006]
　ZHANG Jianhua,LIU Xinke,ZHAO Yan,et al.Automatic detection algorithm for intracranial aneurysm based on 3D-RA image[J].Chinese Journal of Medical Physics,2022,39(8):950-956.[doi:DOI:10.3969/j.issn.1005-202X.2022.08.006]
点击复制

基于3D-RA图像的颅内动脉瘤自动检测算法()
分享到：

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
39卷
期数:
2022年第8期
页码:
950-956
栏目:
医学影像物理
出版日期:
2022-08-04

文章信息/Info

Title:
Automatic detection algorithm for intracranial aneurysm based on 3D-RA image
文章编号:
1005-202X（2022）08-0950-07
作者:
张建华1; 刘新科2; 赵岩1; 杨旭3
1.北京科技大学机械工程学院， 北京 100083； 2.首都医科大学附属北京天坛医院神经介入中心， 北京 100050； 3.河北工业大学机械工程学院， 天津 300401
Author(s):
ZHANG Jianhua1;  LIU Xinke2;  ZHAO Yan1;  YANG Xu1
1. School of Mechanical Engineering, University of Science and Technology Beijing, Beijing 100083, China 2. Neurointerventional Center, Beijing Tiantan Hospital, Capital Medical University, Beijing 100050, China 3. School of Mechanical Engineering, Hebei University of Technology, Tianjin 300401, China
关键词:
颅内动脉瘤; 3D-RA图像; 先验知识; 光流; 卷积神经网络
Keywords:
Keywords: intracranial aneurysm 3D-RA imagepriori knowledge optical flow convolutional neural network
分类号:
R318；TP391.9
DOI:
DOI:10.3969/j.issn.1005-202X.2022.08.006
文献标志码:
A
摘要:
颅内动脉瘤检出率低、破裂后致死率高，是一种严重威胁人类生命健康的高发性脑血管疾病。针对二维卷积神经网络在动脉瘤诊断中对先验知识利用不足问题，基于3D-RA序列图像成像特点，提出一种基于光流可变形卷积的颅内动脉瘤检测算法。采用稠密光流算法获取序列图像之间的光流信息作为先验知识，结合光流信息改进二维卷积计算过程，提出光流可变形卷积模型，从而建立序列图像间的像素级联系。此外，结合光流可变形卷积和标准卷积组成编码模块，实现图像重要特征提取。以北京天坛医院360例临床3D-AR颅内血管造影数据为样本集，测试结果表明：所提方法正确率为0.978 7、精确率为0.983 6、召回率为0.974 7、F1分数为0.979 1、AUC为0.992 4、mAP为0.982 2；与传统网络U-net、Attention U-net相比，该网络对颅内动脉瘤检测更准确；与原有可变形卷积模型相比，光流可变形卷积模型利用光流作为先验知识，提高了网络性能。
Abstract:
Abstract: Intracranial aneurysm which is a kind of high incidence cerebrovascular disease seriously threatening human life and health has low detection rate and high mortality after rupture. Aiming at the problem of insufficient utilization of prior knowledge by 2D convolution neural network in aneurysm diagnosis, an algorithm for intracranial aneurysm detection using deformable convolution integrated with optical flow is proposed based on the imaging characteristics of 3D-RA sequence images. Dense optical flow algorithm is used to obtain the optical flow information between sequential images as prior knowledge.Then the obtained optical flow information is used to improve the 2D convolution calculation process, and an optical flow deformable convolution model is proposed to establish the pixel-level connection between sequential images. In addition, an encoding module is composed of optical flow deformable convolution and standard convolution to extract important features from images. The 3D-AR intracranial angiography data from 360 clinical cases in Beijing Tiantan Hospital are taken as sample set, and the test results showed that the accuracy, precision, recall value, F1 score, AUC and mAP of the proposed method are 0.978 7, 0.983 6, 0.974 7, 0.979 1, 0.992 4 and 0.982 2 respectively. The accuracy of the proposed network in detecting intracranial aneurysm is higher than that of traditional networks U-net and Attention U-net. Compared with traditional deformable convolution model, optical flow deformable convolution model uses optical flow as prior knowledge, which improves network performance.

相似文献/References:

[1]王伟,晁满香,王岚,等.二维数字减影血管造影和三维CT血管成像诊断颅内动脉瘤204例临床分析[J].中国医学物理学杂志,2015,32(05):674.[doi:doi:10.3969/j.issn.1005-202X.2015.05.013]

备注/Memo

备注/Memo:
【收稿日期】2022-02-06 【基金项目】国家自然科学基金（62003128）；天津市自然科学基金（20JCYBJC00610）；河北省自然科学基金（F2020202053） 【作者简介】张建华，博士，教授，研究方向：医疗机器人、医学图像处理，E-mail: eaglezcfx@163.com 【通信作者】赵岩，博士，讲师，研究方向：手术机器人、医学图像处理、深度学习、机器人学习，E-mail: zhaoyan-0312@hotmail.com
更新日期/Last Update: 2022-09-05