This article was originally compiled by the Brainnews World Brain Science Team. Please contact authorization for reprinting.
Imaging technology is increasingly used to generate high-resolution reference maps of brain structure and function. Comparing the experimentally generated brain map with these reference brain maps will help interdisciplinary scientific discoveries. Although recent data sharing plans have increased accessibility to brain maps, data is often shared in different coordinate systems, thus precluding systemic and accurate comparisons. Recently, the Bratislav Misic team introduced their latest research results "neuromaps: structural and functional interpretation of brain maps" in the latest issue of the journal Nature Methods.
This study proposes a neural map, a toolbox for accessing, transforming, and analyzing brain structure and functional annotations. Neuromaps combines open access data with transparent functions for standardization and comparison of brain maps, providing a systematic workflow for the comprehensive structure and functional annotation analysis of the human brain.
Imaging and recording technology is used to generate high-resolution brain maps of the human brain, which provide in-depth understanding of the structure and functional structure of the brain. Such brain maps are increasingly shared on open repositories such as NeuroVault or BALSA, which together provide a comprehensive multimodal perspective on central nervous system . However, these data sharing platforms are limited to surface or volume data and do not integrate standardized analytical workflows. The authors introduce an open access Python toolbox, neuromaps, to enable researchers to systematically share, transform, and compare brain maps (Figure 1). The neuromaps software toolbox is available on https://GitHub.com/netneurolab/neuromaps. On PyPi and Zenodo, it exists as an Docker container. The document can be found on the GitHub page (https://netneurolab.github.io/neuromaps).
Figure 1.neuromaps toolbox function
neuromaps data repository
neuromaps toolbox provides programming access to four standard coordinate system templates: fsaverate, fsLR, CIVET and MNI-152. The neuromaps toolbox also provides access to the brain map repository obtained from published literature (Figure 2). In general, these brain maps of represent more than a decade of human brain mapping research and contain a variety of phenotypes , including the first major component of gene expression , 36 neurotransmitter receptor PET tracer images, glucose and oxygen metabolism, cerebral blood flow and capacity, cortical thickness, T1 weighted/T2 weighted MRI ratio, six typical MEG bands, intrinsic time scale, evolutionary expansion, three developmental expansion maps, the first 10 functional connection gradients, intersubjective variability and the first principal component cognitive activation derived from NeuroSynth. This data repository is organized by tags and can be downloaded directly from neuromaps.
Figure 2. Brain map from published literature
Transformation between coordinate systems
Transformation between volume-based and surface-based coordinate systems depends on the registration fusion framework (Figure 3a), while transformation between surface-based coordinate systems uses a multimodal surface matching (MSM) framework (Figure 3b). By default, neuromaps returns brain maps in low-resolution map space, which ensures that neuromaps do not artificially create upsampled data. Overall, the neuromaps toolbox implements robust conversion between coordinate systems to facilitate standardization of the neuroimaging workflow (Figure 3c,d).
Figure 3. Transformation between coordinate systems
Demonstration neural diagram toolbox
In order to show the practicality of neuromaps, the author applied three independent analysis workflows.First, the authors applied the neuromaps toolbox to a volumetric map of cortical thinning derived from a T1-weighted MRI scan from 133 schizophrenia patients with n = 113 controls from Northwestern University schizophrenia data and software (Figure 4a).
Figure 4. Using neuromaps
Next, the authors applied the same analysis workflow to a surface-based evolutionary extension brain map, which represents the cortical surface area expansion from macaque to humans (Figure 4b). Finally, the authors analyzed 20 brain map samples from published literature over the past decade (2011-2021), including two microstructures, four metabolisms, three functions, four extensions, six band-specific electrophysiological signaling power, and one genomic map. These graphs are then transformed from their original representations to a space defined by each of the four standard coordinate systems, with seven different representations in total (Figure 2). Finally, the authors calculated pairwise correlations between all maps in each system and evaluated the statistical significance of these relationships using a spatial zero model (Fig. 5).
Figure 5. Application of neuromaps in 20 brain maps
Summary
This article introduces an open source Python package, neuromaps, used for human brain research . As the field continues to generate new brain maps, the authors hope that neuromaps will provide researchers with a standardized set of workflows to better understand what this data can tell us about the human brain. As researchers adopt the neuromaps toolbox, users can add comments from emerging technologies and datasets. This will enable the brain map to systematic context for multiple normative annotations from different data types and disciplines, resulting in standardized results reports and inspire for subsequent work. neuromaps is a step towards a comprehensive analysis of multi-modal, multi-scale neuroscience .
References
Markello, R.D., Hansen, J.Y., Liu, ZQ. et al. neuromaps: structural and functional interpretation of brain maps. Nat Methods (2022). https://doi.org/10.1038/s41592-022-01625-w
Compiled by: Ayden (brainnews creative team)
Reviewed: Simon (brainnews editorial department)
