Harmonization-of-CT-Convolution-Kernels

Two models used for hrmonization of CT reconstruction kernels:

Model CNNa was applied to convert reconstruction kernels from sharp (ie,Siemens B70f, GE Lung amd Edge) to smooth (ie, Siemens B30f). Model CNNb was applied to convert reconstruction kernels from smooth (ie, Siemens B30f, GE Standard, Soft) to sharp (ie, Siemens B70f).

Environments

ubuntu 14.04 LTS, python 3.7, Pytorch framework.