基于深度学习的图像重建算法在CT图像重建中的应用进展  

Application of Image Reconstruction Algorithm Based on Deep Learning in CT Image

作　　者：王梅 花立春 唐颖 边传振 WANG Mei;HUA Lichun;TANG Ying;BIAN Chuanzhen(Department of Radiology,Children’s Hospital of Nanjing Medical University,Nanjing Jiangsu 210008,China)

机构地区：[1]南京医科大学附属儿童医院放射科,江苏南京210008

出　　处：《中国医疗设备》2022年第12期157-161,共5页China Medical Devices

基　　金：江苏省妇幼保健协会科研项目(FYX201918)。

摘　　要：随着CT扫描技术的飞速发展,在人工智能大背景下,深度学习图像重建(Deep Learning Image Reconstruction,DLIR)算法在CT成像中得到了广泛的应用,相对于目前常用的图像重建算法,该算法能够保证低辐射剂量条件下的图像细节显示,并在一定程度上降低图像噪声,且还能够减少图像重建时间,从而使低辐射剂量扫描成为可能。本文对DLIR算法在人体不同部位CT检查的应用现状及研究进展进行综述,以期进一步推广DLIR算法,在降低公众CT检查辐射剂量的同时改善图像质量。With the rapid development of CT scanning technology,in the context of artificial intelligence,the deep learning image reconstruction(DLIR)algorithm has been widely used in CT imaging.Compared with the current commonly used image reconstruction algorithm,this algorithm can ensure the display of image details under low radiation dose conditions,reduce image noise to a certain extent,and reduce the image reconstruction time.This paper reviewed the application status and research progress of DLIR algorithm in CT examination of different parts of the human body,hoping that DLIR algorithm can be further promoted to reduce the radiation dose of public CT examination and improve the image quality.

关 键 词：深度学习 图像重建 辐射剂量 计算机断层扫描 

分 类 号：R445.2[医药卫生—影像医学与核医学] TP391.41[医药卫生—诊断学]