DDR-Net: Dividing and Downsampling Mixed Network for Diffeomorphic Image Registration
This work addresses memory constraints in medical image registration, which is crucial for accurate 3D volume alignment in healthcare applications, representing an incremental improvement over existing methods.
The paper tackles the memory limitations in deep diffeomorphic registration for high-dimensional 3D medical images by proposing DDR-Net, which preserves image information at multiple scales through a mixed architecture of downsampling and dividing input images, and it outperforms existing approaches on three public datasets.
Deep diffeomorphic registration faces significant challenges for high-dimensional images, especially in terms of memory limits. Existing approaches either downsample original images, or approximate underlying transformations, or reduce model size. The information loss during the approximation or insufficient model capacity is a hindrance to the registration accuracy for high-dimensional images, e.g., 3D medical volumes. In this paper, we propose a Dividing and Downsampling mixed Registration network (DDR-Net), a general architecture that preserves most of the image information at multiple scales. DDR-Net leverages the global context via downsampling the input and utilizes the local details from divided chunks of the input images. This design reduces the network input size and its memory cost; meanwhile, by fusing global and local information, DDR-Net obtains both coarse-level and fine-level alignments in the final deformation fields. We evaluate DDR-Net on three public datasets, i.e., OASIS, IBSR18, and 3DIRCADB-01, and the experimental results demonstrate our approach outperforms existing approaches.