ROIMAPer: An Open Source Framework for Rapid and Accurate Atlas Based Registration of Individual Brain Images in FIJI

Abstract

The brains remarkable complexity and cellular heterogeneity necessitate precise anatomical annotation to ensure that imaging-based analyses accurately resolve region-specific features. Few computational tools currently exist that allow for the accurate and rapid registration of single brain images to standard brain atlases. To address this limitation, we developed ROIMAPer, a novel FIJI plugin for rapid registration of individual brain slices. ROIMAPer includes eight atlases spanning mouse, rat, and human brain anatomy across multiple developmental stages, making it broadly applicable across diverse experimental contexts. It allows for linear and affine scaling of the reference atlas to the experimental image and is optimized for serial processing of large quantities of images. We demonstrated the accuracy of ROIMAPer through quantification of in situ hybridization data from the Allen Gene Expression Atlas of seven marker genes across major brain regions and of four marker genes across hippocampal subfields. Quantification of marker genes within their assigned brain regions closely matched the ground truth across all major regions. At a finer resolution, marker-gene quantification within hippocampal subregions aligned with the experimental data, although discrepancies with the ground truth were observed for Mcu. Overall, ROIMAPer provides broad utility for open-source brain image analysis from multiple species. Significance StatementWe present an open-source, user-friendly, and accessible tool for registration of individual brain slices to anatomical reference atlases, compatible with the image analysis platform FIJI. The field lacks tools that offer a span of cross-species atlases, FIJI-compatibility, intuitive linear scaling methods, and low user-input without requiring high computational skill. Our tool minimizes user-involvement, allows for processing of larger datasets through more effective resource management, and speeds-up previously tedious processing steps.