A Novel Non-External Marker 4D-CT Sorting Algorithm

This article describes a full automated non-external marker 4D-CT sorting algorithm which produce coherent 4D-CT volume bins. Firstly, a normalized cross correlation (NCC) sorting algorithm is applied to get initial 4D-CT data sets. Then a spatial continuity correction method is applied to optimize the 4D-CT data sets. The spatial continuity correction method is based on 2D image free-form registration. It can find the discontinuities and "flatness" in CT volume and correct them by using greedy algorithm. Two 4D-CT data sets sorted by spatial continuity correction algorithm show significant improvement compared with that sorted by NCC algorithm only. The other two sets have not obvious improvement because NCC algorithm already sorts them very well. The algorithm described in this article can remove artifacts and produce more coherent 4D-CT data sets without any external marker.