Prediction of cervical cancer recurrence using textural features extracted from 18F-FDG PET images acquired with different scanners

This study showed that radiomic features could predict local recurrence of LACC better than SUVmax. Further investigation is needed before applying a model designed using data from one PET scanner to another.