基于深度学习的电阻抗断层成像图像重建算法研究综述  

Reconstruction algorithm in electrical impedance tomography based on deep learning

作　　者：张里园 李磊 刘学超 付峰 金莉[2] 杨滨 ZHANG Li-yuan;LI Lei;LIU Xue-chao;FU Feng;JIN Li;YANG Bin(Military Biomedical Engineering School,Air Force Medical University,Xi'an 710032,China;The School of Software Engineering,Xi'an Jiaotong University,Xi'an 710049,China)

机构地区：[1]空军军医大学军事生物医学工程学系,西安710032 [2]西安交通大学软件学院,西安710049

出　　处：《医疗卫生装备》2024年第4期98-103,共6页Chinese Medical Equipment Journal

基　　金：国家自然科学基金面上项目(52377230);国家自然科学基金重点项目(51837011)。

摘　　要：介绍了电阻抗断层成像(electrical impedance tomography,EIT)传统图像重建算法的局限性和深度学习在非线性图像重建中的优点,综述了基于深度学习的直接重建法、间接重建法和隐式重建法3种EIT图像重建算法的研究进展,分析了基于深度学习的EIT图像重建算法存在的不足,指出了基于深度学习的EIT图像重建算法应在提高数据集质量、增强图像可解释性以及改善成像分辨力等方面实现进一步的技术突破。The limitations of traditional image reconstruction algorithms of electrical impedance tomography(EIT)and the advantages of deep learning in nonlinear image reconstruction were introduced.The research progress of three EIT image reconstruction algorithms based on deep learning was summarized,including direct reconstruction method,indirect recon-struction method and implicit reconstruction method.The deficiencies of EIT image reconstruction algorithms based on deep learning were analyzed,and it's pointed out that EIT image reconstruction algorithms based on deep learning should realize further technical breakthroughs in increasing the quality of datasets,enhancing the image interpretability and improving the imaging resolving power.

关 键 词：深度学习 电阻抗断层成像 图像重建 图像重建算法 

分 类 号：R318[医药卫生—生物医学工程] R319[医药卫生—基础医学]