摘要
电阻抗断层成像(EIT)技术在无损伤、装置简单等方面有着巨大优势,但其存在重构结果质量不理想问题。鉴于此,利用深度学习方法建立基于电阻抗断层成像的映射网络模型并进行比较分析。结果显示,在结构相似性上,所提算法高于0.8,而比较算法均低于0.6,表明所提模型在图像重构质量具有优势,具备一定投用价值。
Electrical impedance tomography(EIT)technology has enormous advantages in terms of non-destructive and simple installation,but the quality of reconstruction results is not ided.In view of this,a mapping network model based on electrical impedance tomography is established by deep learning method and compared.The results show that in terms of structural similarity,the proposed algorithm is higher than 0.8,while the comparative algorithms are all lower than 0.6,which indicates that the proposed model has advantages in image reconstruction quality and has certain practical value.
作者
张恒
任晓康
徐金伍
李梦
谢利德
ZHANG Heng;REN Xiaokang;XU Jinwu;LI Meng;XIE Lide(Biomedical Engineering Department,Chengde Medical University,Chengde 053000,China;Peking University Third Hospital,Beijing 100191,China)
出处
《微型电脑应用》
2024年第7期106-108,113,共4页
Microcomputer Applications
关键词
EIT
原穴
电阻抗
图像重建
DNN
EIT
original acupoint
electrical impedance
image reconstruction
DNN