RheMAP-Surf

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

---

The RheMAP-Surf project provides precomputed deformation spheres for performing surface-based transformations among several widely used rhesus macaque brain templates, including NMT v2.0 (symmetric and asymmetric), D99, MEBRAINS v2.0, and Yerkes19. These transformations were generated using FSL's MSM (Multimodal Surface Matching).

This repository includes:

• Deformation spheres
• Standardized spherical and anatomical surfaces for each template
• Source template files
• A demo with detailed instructions and example code
• Accompanying documentation covering processing methods, alignment quality assessment, and visualization

In addition, the code and intermediate results produced during the generation of these surface alignments are available via Zenodo:

Authors: Zi-Xuan Zhou & Ting Xu (Child Mind Institute)

Citation: Zhou, Z.-X., & Xu, T. (2025). RheMAP-Surf: Surface Transformations Among Commonly Used Rhesus Macaque Brain Templates (1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.15642802

---

Introduction

In macaque neuroimaging, the coexistence of multiple templates poses challenges for cross-study comparability, data integration, and reproducibility. The RheMAP project addresses this issue at the volume level by providing ANTs-derived nonlinear volume registrations among commonly used rhesus macaque templates. However, these volume registrations do not directly resolve discrepancies across surface spaces.

To date, there is no community standard for inter-template surface transformations in macaques. Different alignment protocols can lead to substantial variability, complicating comparisons across studies. Moreover, it remains unclear how closely surface-based transformations replicate the effects of volume-based warps, and whether systematic differences exist. The RheMAP-Surf project aims to fill these gaps by providing surface-based counterparts to the RheMAP volume registrations and evaluating their correspondence.

---

Project Overview

RheMAP-Surf provides precomputed surface deformation fields across five macaque brain templates:

• NMTv2.0-sym (symmetric NMT v2.0)
• NMTv2.0-asym (asymmetric NMT v2.0)
• D99
• MEBRAINS (MEBRAINS v2.0)
• YRK (Yerkes19)

We standardize the naming of their spherical and anatomical surfaces and corresponding deformation spheres, all of which follow consistent conventions. See the demo/ for usage examples.

Since deformation spheres from F99 to Yerkes19 have previously been released, we do not include F99-based transformations to avoid introducing additional variability. Users may use Yerkes19 as a bridge, via the F99-to-Yerkes19 transformations available in templates/YRK/.

We provide two alternative sets of deformation spheres:

• MSM-curv: conventional MSM alignment using FreeSurfer-derived mean curvature as the driving feature. This method reflects a common practice in current neuroimaging studies.
• MSM-aprx: MSM alignment based on RheMAP-projected random features, designed to approximate ANTs-derived volume registrations and bridge surface and volume transformations.

Detailed quality assessments and visualizations for both sets, as well as comparison analyses between them, are provided in the documents below.

---

Documentation

For processing workflows, alignment strategies, and evaluation results, please refer to the following markdown reports:

1. preprocessing.md
   Surface preprocessing and medial wall harmonization

2. msm_alignment_with_mean_curvature.md
   Surface alignment using mean curvature (MSM-curv)

3. msm_alignment_with_projected_features.md
   Surface alignment using RheMAP-projected features (MSM-aprx)

4. msm_alignment_difference.md
   Comparison between MSM-curv and MSM-aprx alignment results

---

If you use this resource, please cite the Zenodo archive () and acknowledge the RheMAP-Surf project.

---