正规化(语言学)
电阻抗断层成像
算法
共轭梯度法
全变差去噪
相关系数
水肿
肺水肿
断层摄影术
数学
肺
计算机科学
医学
人工智能
图像(数学)
放射科
内科学
统计
作者
Qi Wang,Zepeng Hao,Xiuyan Li,Xiaojie Duan,Jianming Wang
标识
DOI:10.1109/icfeict59519.2023.00028
摘要
In order to continuously and visually monitor the process of pulmonary edema, electrical impedance tomography (EIT) is used to monitor the dynamic changes of pulmonary edema in real time through the distribution of regional electrical parameters of the lung, and a 3D-TV regularized three-dimensional EIT imaging method based on Split-Bregman algorithm is proposed. Compare and evaluate the relative error (RE), correlation coefficient (CC) and structural similarity index measurement (SSIM) of the reconstructed image of 3D-TV regularization method with traditional conjugate gradient (CG) algorithm and total variation regularization (TV) algorithm. Finally, the 3D-TV regularization method was applied to the rat pulmonary edema experiment to monitor and analyze the whole process of rat pulmonary edema. The results show that compared with the traditional TV algorithm, the 3D-TV regularization method reduces the RE of 3D EIT imaging by 9.1%, CC elevated by 9.1%, and SSIM elevated by 7.7%, indicating that the proposed 3D-TV has greater advantages for 3D EIT image reconstruction, and can monitor and image the whole process of pulmonary edema in rats.
科研通智能强力驱动
Strongly Powered by AbleSci AI