zjk1988 / DAN-Net

A Pytorch implementation of Dual-domain Adaptive-scaling Non-local Network for CT Metal Artifact Reduction.
18 stars 3 forks source link

DAN-Net


paper:DAN-Net: Dual-Domain Adaptive-Scaling Non-local Network for CT Metal Artifact Reduction

by Tao Wang scuer_wt@scu.stu.edu.cn. This repository implements DAN-Net by Pytorch for metal artifacts reduction (MAR) in CT images. We re-implement the reconstruction layer according to the formulas in DuDoNet and we used Numba in Python to achieve parallel computing.

This work was accepted by MICCAI2021, and the extended version was accepted by Physics in Medicine & Biology (PMB).

Prerequisites

This repository needs the following system settings: