The "hidden noise" problem in MR image reconstruction.

Published in: Magnetic resonance in medicine (2024)

Reference data containing hidden noise can substantially mislead the ranking of image reconstruction methods when using conventional error metrics, but this issue can be mitigated with alternative error metrics.

Keyphrases

deep learning
air pollution
electronic health record
magnetic resonance
big data
magnetic resonance imaging
artificial intelligence
computed tomography