基于生成对抗网络的低剂量CT图像去噪方法研究

Low-Dose CT Denoising Based on Generative Adversarial Network

计算机断层技术在医学临床诊断上发挥着十分重要的作用,但是在带来极大方便的同时,CT扫描时存在的辐射也严重威胁人类的健康。随着人们对CT图像认识的不断加深,尽可能减少受检者的辐射剂量已成为当今影像学重要的研究方向,然而降低辐射量所重建出的CT图像会出现噪声、伪影,严重影响临床诊断。本文基于生成对抗网络,对重建后的低剂量CT图像进行去噪处理。实验结果证明,本文所采用的基于生成对抗网络的低剂量CT图像去噪方法,可以有效消除噪声与伪影,并且可以保留更多的细节信息以帮助临床诊断。

Computer tomography technology plays an important role in medical clinical diagnosis.While bringing great convenience,the radiation presented in CT scans seriously threatens the health of human.As people’s understanding of CT images,reducing the radiation dose as much as possible has become an important research in medical imaging.However,there still exist noise in the CT images reconstructed by reducing the radiation dose,which seriously affects clinical diagnosis.This paper proposed a novel method for low-dose CT denoising based on generative adversarial network.The experimental results of the proposed method show that our method can effectively eliminate noise and artifacts,and maintain more detail information to help clinical diagnosis.