MRI Image Reconstruction via Learning Optimization Using Neural ODEs
This work addresses MRI reconstruction for medical imaging, presenting a novel approach with incremental improvements in performance and efficiency.
The authors tackled MRI image reconstruction by modeling it as an optimization problem using neural ordinary differential equations (ODEs), achieving better reconstruction results and greater parameter efficiency compared to methods like UNet and cascaded CNN.
We propose to formulate MRI image reconstruction as an optimization problem and model the optimization trajectory as a dynamic process using ordinary differential equations (ODEs). We model the dynamics in ODE with a neural network and solve the desired ODE with the off-the-shelf (fixed) solver to obtain reconstructed images. We extend this model and incorporate the knowledge of off-the-shelf ODE solvers into the network design (learned solvers). We investigate several models based on three ODE solvers and compare models with fixed solvers and learned solvers. Our models achieve better reconstruction results and are more parameter efficient than other popular methods such as UNet and cascaded CNN. We introduce a new way of tackling the MRI reconstruction problem by modeling the continuous optimization dynamics using neural ODEs.