Implicit neural representations for joint decomposition and registration of gene expression images in the marmoset brain
This work addresses the challenge of accurately quantifying gene expression in brain images for neuroscience research, but it is incremental as it builds on existing implicit neural representation methods.
The authors tackled the problem of registering gene expression images with varying patterns to a brain template by proposing a method that jointly performs registration and decomposition using implicit neural representations and an image exclusion loss. Their approach outperformed other registration techniques on 2D marmoset brain images.
We propose a novel image registration method based on implicit neural representations that addresses the challenging problem of registering a pair of brain images with similar anatomical structures, but where one image contains additional features or artifacts that are not present in the other image. To demonstrate its effectiveness, we use 2D microscopy $\textit{in situ}$ hybridization gene expression images of the marmoset brain. Accurately quantifying gene expression requires image registration to a brain template, which is difficult due to the diversity of patterns causing variations in visible anatomical brain structures. Our approach uses implicit networks in combination with an image exclusion loss to jointly perform the registration and decompose the image into a support and residual image. The support image aligns well with the template, while the residual image captures individual image characteristics that diverge from the template. In experiments, our method provided excellent results and outperformed other registration techniques.