RheMAP-Surf: Surface Transformations Among Commonly Used Rhesus Macaque Brain Templates

General description This archive contains all input data, outputs, and code used in the RheMAP-Surf project, which provides precomputed surface-based transformations among widely used rhesus macaque brain templates, including NMT v2.0 (symmetric and asymmetric), D99, MEBRAINS v2.0, and Yerkes19. These transformations are implemented as deformation spheres compatible with surface-based pipelines. Deformation fields were derived using FSL’s MSM (Multimodal Surface Matching), via two distinct alignment strategies: MSM-curv: Conventional surface alignment using FreeSurfer-derived mean curvature as the driving feature. MSM-aprx: A novel method that uses ANTs-derived RheMAP volume warps to project structured random features to the surface, guiding alignment that approximates volume transformations across surface spaces. All deformation outputs are organized by template pairs and hemisphere, and evaluated through both visual quality control and quantitative metrics. Rationale In macaque neuroimaging, the use of multiple anatomical templates has led to fragmented practices, hindering cross-study comparability and data integration. While the RheMAP project provides nonlinear ANTs-derived volume registrations to bridge differences across template spaces, these registrations operate in volume space and do not directly account for differences across surface spaces. Currently, there is no standardized framework for inter-template surface transformations in macaques. Surface alignment methods vary in implementation and feature choices, potentially introducing methodological variability. It also remains unclear how well surface-based transformations capture or approximate the effects of volume-based warps. The RheMAP-Surf project fills these gaps by generating surface-to-surface deformation spheres across widely used macaque templates and systematically evaluating their correspondence to RheMAP volume registrations. Code and usage All shell scripts and supporting Python functions used to generate the transformations are included in the code/ directory. A complete example workflow demonstrating usage of the deformation spheres is provided in the demo/ folder. Data structure The archive includes: templates/: External template data used as inputs. Original filenames are preserved for traceability. surfaces/: Final standardized anatomical and spherical surfaces for each template. deformation_spheres/: Precomputed deformation spheres for both alignment methods (MSM-curv and MSM-aprx). demo/: Instructions and example code for using the deformation spheres. code/: All preprocessing, alignment, and evaluation scripts, including intermediate steps. data/: Intermediate results (e.g., curvature features, medial wall masks, MSM alignment outputs, and evaluation files). figures/: Visual outputs generated for evaluation and documentation. Note Documentation presenting processing details and visual outputs is available on the RheMAP-Surf GitHub page (not included in this archive). While deformation spheres for F99 template are not provided, Yerkes19 can be used as a bridge template if needed (see templates/YRK/). For questions, feedback, or contributions, please use the GitHub Issues page. If you use the deformation spheres, any part of this dataset, or are inspired by the RheMAP-Surf project, please cite this Zenodo archive.