users > RE: How do we know whether the registration result is good
Feb 1, 2023 11:02 AM | Sandor Kovacs
RE: How do we know whether the registration result is good
As a first step you can evaluate the results visually and critically, The registered images should visually align well, with similar structures in the images appearing to overlap and the registered images should match the expected anatomy. Then apply image similarity metrics such as mean squared difference (MSD), normalized cross-correlation (NCC), mutual information (MI), structural similarity index (SSIM), should have values close to 1, indicating a high degree of similarity between the registered images.
|Ziqi Song||Feb 1, 2023|
|Sandor Kovacs||Feb 1, 2023|