Deformation-Compensated Learning for Image Reconstruction Without Ground Truth

Weijie Gan, Yu Sun, Cihat Eldeniz, Jiaming Liu, Hongyu An, Ulugbek S. Kamilov

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Deep neural networks for medical image reconstruction are traditionally trained using high-quality ground-truth images as training targets. Recent work on Noise2Noise (N2N) has shown the potential of using multiple noisy measurements of the same object as an alternative to having a ground-truth. However, existing N2N-based methods are not suitable for learning from the measurements of an object undergoing nonrigid deformation. This paper addresses this issue by proposing the deformation-compensated learning (DeCoLearn) method for training deep reconstruction networks by compensating for object deformations. A key component of DeCoLearn is a deep registration module, which is jointly trained with the deep reconstruction network without any ground-truth supervision. We validate DeCoLearn on both simulated and experimentally collected magnetic resonance imaging (MRI) data and show that it significantly improves imaging quality.

Original languageEnglish
Pages (from-to)2371-2384
Number of pages14
JournalIEEE Transactions on Medical Imaging
Issue number9
StatePublished - Sep 1 2022


  • Inverse problems
  • deep learning
  • image reconstruction
  • magnetic resonance imaging (MRI)


Dive into the research topics of 'Deformation-Compensated Learning for Image Reconstruction Without Ground Truth'. Together they form a unique fingerprint.

Cite this