一种基于Fractal-DenseNet的电磁血管断层图像重建算法

doi:10.12068/j.issn.1005-3026.2023.07.003

东北大学学报（自然科学版） ›› 2023, Vol. 44 ›› Issue (7): 931-937.DOI: 10.12068/j.issn.1005-3026.2023.07.003

一种基于Fractal-DenseNet的电磁血管断层图像重建算法

杨丹^1,2，王雨佳¹，辛采凝¹，徐彬³

(1.东北大学信息科学与工程学院，辽宁沈阳110819； 2.东北大学智能工业数据解析与优化教育部重点实验室，辽宁沈阳110819； 3.东北大学计算机科学与工程学院，辽宁沈阳110169)

发布日期:2023-07-13
通讯作者: 杨丹
作者简介:杨丹(1979-)，女，辽宁营口人，东北大学副教授.
基金资助:
国家自然科学基金资助项目(U22A20221)；辽宁省自然科学基金资助项目(2021-MS093)；辽宁省教育厅基础科学研究项目(LJKZ0014).

A Reconstruction Algorithm for Electromagnetic Vascular Tomography Image Based on Fractal-DenseNet

YANG Dan^1,2， WANG Yu-jia¹， XIN Cai-ning¹， XU Bin³

1. School of Information Science & Engineering， Northeastern University， Shenyang 110819， China; 2. Key Laboratory of Data Analytics & Optimization for Smart Industry， Northeastern University， Shenyang 110819， China; 3. School of Computer Science & Engineering， Northeastern University， Shenyang 110169， China.

Published:2023-07-13
Contact: XU Bin
About author:-
Supported by:
-

摘要/Abstract

摘要： 针对电磁血管断层图像重建中存在的欠定性、病态性，提出一种基于Fractal-DenseNet的电磁血管断层图像重建算法.基于血流磁电耦合效应的血管断层图像重建原理，将FractalNet的分形思想与DenseNet的密集紧密连接思想相结合，构建了一种适用于反演血液流速分布的Fractal-DenseNet网络模型，用于血管断层图像重建.通过人体前臂尺动脉血流磁电耦合测量模型，建立血管断层剖面流速和血流磁电效应引起的电压信号的对应数据对，分别作为网络模型输入和输出；通过监督学习，实现基于血管断层流速分布的血管断层图像重建.结果表明:Fractal-DenseNet重建结果的均方根误差和相关系数分别为0.0078，99.28%；本文算法具有良好的抗噪性，可在复杂边界条件下实现血管断层图像重建.

关键词: 血管断层图像重建；Fractal-DenseNet；血流磁电耦合效应；人体前臂尺动脉；监督学习

Abstract: In view of the lack of underdetermination and pathology in electromagnetic vascular tomography image reconstruction， a Fractal-DenseNet based vascular tomography image reconstruction algorithm is proposed. With the principle of vascular tomography image reconstruction based on the magneto-electric coupling effect of blood flow， combining FractalNet’s fractal idea with DenseNet’s dense connection idea， a Fractal-DenseNet network model suitable for the inversion of blood flow velocity distribution was built for the reconstruction of vascular tomography image. Based on the magneto-electric coupling measurement model of human forearm ulnar artery flow， the corresponding data pairs of vessel section velocity and voltage signal caused by magneto-electric effect of blood flow were established， which were respectively used as the input and the output of the network model. Through supervised learning， the image reconstruction based on the flow velocity distribution of the vessel is realized. The results show that the root mean square error and correlation coefficients of Fractal-DenseNet reconstruction are 0.0078 and 99.28%， respectively. The proposed model has good anti-noise performance and can be used to reconstruct vascular tomography images under complex boundary conditions.

Key words: vascular tomography image reconstruction; Fractal-DenseNet; magneto-electric coupling effect of blood flow; ulnar artery of human forearm; supervised learning

中图分类号:

TH776

杨丹，王雨佳，辛采凝，徐彬. 一种基于Fractal-DenseNet的电磁血管断层图像重建算法[J]. 东北大学学报（自然科学版）, 2023, 44(7): 931-937.

YANG Dan， WANG Yu-jia， XIN Cai-ning， XU Bin. A Reconstruction Algorithm for Electromagnetic Vascular Tomography Image Based on Fractal-DenseNet[J]. Journal of Northeastern University(Natural Science), 2023, 44(7): 931-937.

参考文献

[1]Ali M.Development of an electromagnetic induction method for non-invasive blood flow measurement［D］.Huddersfield:University of Huddersfield，2016.
[2]Yang D，Liu Y，Xu B，et al.A blood flow volume linear inversion model based on electromagnetic sensor for predicting the rate of arterial stenosis ［J］.Sensors，2019，19(13): 3006.
[3]Wang J，Han B.Application of a class of iterative algorithms and their accelerations to Jacobian-based linearized EIT image reconstruction［J］.Inverse Problems in Science and Engineering，2021，29(8): 1108-1126.
[4]Harikumar R，Prabu R，Raghavan S.Electrical impedance tomography (EIT) and its medical applications:a review［J］.The International Journal of Soft Computing and Engineering，2013，3(4):193-198.
[5]Wei H Y，Soleimani M.Four dimensional reconstruction using magnetic induction tomography:experimental study［J］.Progress in Electromagnetics Research，2012，129:17-32.
[6]Chen D，Davies M E.Deep decomposition learning for inverse imaging problems［C］//European Conference on Computer Vision.Glasgow，2020:510-526.
[7]Wu X J，Xu M D，Li C D，et al.Research on image reconstruction algorithms based on autoencoder neural network of restricted boltzmann machine (RBM)［J］.Flow Measurement and Instrumentation，2021，80:102009.
[8]Huang S W，Cheng H M，Lin S F.Improved imaging resolution of electrical impedance tomography using artificial neural networks for image reconstruction［C］//2019 41^st Annual International Conference of the IEEE Engineering in Medicine and Biology Society.Berlin，2019:1551-1554.
[9]Yoo J，Sabir S，Heo D，et al.Deep learning diffuse optical tomography［J］.IEEE Transactions on Medical Imagin，2019，39(4):877-887.
[10]Liu Y，Yang D，Duo Y，et al.Numerical model and finite element simulation of arterial blood flow profile reconstruction in a uniform magnetic field［J］.Journal of Physics D:Applied Physics，2020，53(19):195402.
[11]Li J Y，Wen H B，Zhang K H，et al.Research progress in electromagnetic interference suppression of magnetically coupled radio energy transmission systems［J］.Proceedings of the CSEE，2022，42(20):7387-7403.
[12]Larsson G，Maire M，Shakhnarovich G.FractalNet:ultra-deep neural networks without residuals［J］.arXiv preprint arXiv:1605.07648，2016.
[13]Huang G，Liu Z，van Der Maaten L，et al.Densely connected convolutional networks［C］//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Honolulu，2017:4700-4708.

一种基于Fractal-DenseNet的电磁血管断层图像重建算法

A Reconstruction Algorithm for Electromagnetic Vascular Tomography Image Based on Fractal-DenseNet

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics

本文评价