Robustness of PET Radiomics Features: Impact of Co-Registration with MRI

Radiomics holds great promise in the field of cancer management. However, clinical application radiomics has been hampered by uncertainty about robustness features extracted from images. Previous studies have reported that are sensitive to changes voxel size resampling and interpolation, image perturbation, or slice thickness. This study aims observe variability positron emission tomography (PET) under impact co-registration with magnetic resonance imaging (MRI) using difference percentage coefficient, Spearman’s correlation coefficient for three groups images: (i) original PET, (ii) PET after T1-weighted MRI (iii) FLAIR MRI. Specifically, seventeen patients brain cancers undergoing [11C]-Methionine were considered. Successively, images co-registered sequences 107 each mentioned group The analysis revealed shape features, first-order two subgroups higher-order possessed a good robustness, unlike remaining which showed large differences coefficient. Furthermore, approximately 40% selected differed is an important consideration users conducting constraints avoid errors diagnosis, prognosis, outcome prediction.