Since the introduction of CT technology,the diagnosis and treatment of diseases have been unprecedentedly developed. At the same time,the radiation dose generated during CT examination is a hotspot of concern for doctors and patients,so how to minimize the radiation dose under the premise of ensuring image quality has become the focus of CT research. Deep learning image reconstruction (DLIR) algorithm is an emerging CT reconstruction technique based on convolutional neural network,which can reduce the radiation dose and contrast dosage and optimize the image quality compared with the traditional filtered back projection,iterative reconstruction algorithms. Therefore,this paper reports a general overview of the technological evolution of the DLIR algorithm,its basic principles,technical advantages and clinical applications.
关键词
深度学习/图像重建算法/计算机体层成像/图像质量
Key words
Deep learning/Image reconstruction algorithm/Computed tomography/Image quality
维普
万方数据