Deformation-Compensated Learning for Image Reconstruction without Ground Truth
This addresses a challenge in medical imaging for researchers and practitioners by enabling reconstruction from deformed measurements without ground truth, though it is incremental as it builds on Noise2Noise methods.
The paper tackled the problem of training deep neural networks for medical image reconstruction without ground-truth images, especially when objects undergo nonrigid deformation, by proposing the DeCoLearn method that jointly trains a deep registration module with a reconstruction network, resulting in significantly improved imaging quality on simulated and experimental MRI data.
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.