Artificial CT Data Generation Method with Known Ground-Truth for Image Registration with Missing Data
Marek Wodziński, Andrzej Skalski · 2018
In this work we propose a method to generate artificial data for evaluation of image registration algorithms dedicated to the missing data problem. The generated data can be used as an alternative to the evaluation based on the artificial deformations which resemble the missing data creation process. The proposed method is dedicated to the initial evaluation, before using mechanical phantoms and real, medical data. We describe step-by-step the proposed method and compare the deformation fields calculated between artificially generated data and data deformed using the artificial deformation fields. The main advantage of the proposed method, compared to the known deformations, is the possibility to model a full surgical resection.