Automatic motion artefact detection in brain T1-weighted magnetic resonance images from a clinical data warehouse using synthetic data

Sophie Loizillon, Simona Bottani, Aurélien Maire, Sebastian Ströer, Didier Dormont, Olivier Colliot, Ninon Burgos

Abstract: Highlights•We propose a novel transfer learning framework from research to clinical data for the automatic detection of motion in 3D T1w brain MRI.•We generate synthetic motion in MR images of research databases to train a motion CNN classifier.•We generalise our model to clinical data with an effective fine-tuning technic.•We perform a very large-scale validation on 500 clinical MRIs of the AP-HP CDW with manual annotations regarding motion.