Deep learning based registration using spatial gradients and noisy segmentation labels
This work addresses image registration in medical analysis, but it is incremental as it builds on existing deep learning approaches and reports competitive results in a specific challenge.
The paper tackles medical image registration by proposing a deep learning method with symmetric formulation and integration of public datasets, achieving a mean dice of 0.64 for task 3 and 0.85 for task 4 in the Learn2Reg challenge.
Image registration is one of the most challenging problems in medical image analysis. In the recent years, deep learning based approaches became quite popular, providing fast and performing registration strategies. In this short paper, we summarise our work presented on Learn2Reg challenge 2020. The main contributions of our work rely on (i) a symmetric formulation, predicting the transformations from source to target and from target to source simultaneously, enforcing the trained representations to be similar and (ii) integration of variety of publicly available datasets used both for pretraining and for augmenting segmentation labels. Our method reports a mean dice of $0.64$ for task 3 and $0.85$ for task 4 on the test sets, taking third place on the challenge. Our code and models are publicly available at https://github.com/TheoEst/abdominal_registration and \https://github.com/TheoEst/hippocampus_registration.