TL;DR
This study uses a novel computational neuroanatomy framework to disentangle how brain evolution reshapes cortical areas and their connections. By aligning primate brains based on myelin maps, they show that for most white matter tracts, evolutionary changes in brain area size and position explain observed differences in connectivity. However, the arcuate fasciculus, crucial for language, shows unique evolutionary adaptations beyond simple relocation.
What Problem Does This Paper Address?
The human brain exhibits remarkable specializations, particularly in the temporal lobe, which are thought to underpin complex behaviors like language. Understanding how these specializations evolved is a major challenge. Two main hypotheses exist: 1) local expansion and relocation of existing brain areas, and 2) changes in the connections (white matter tracts) between these areas. It's unclear to what extent these two mechanisms operate independently or interactively. This paper aims to disentangle these two evolutionary processes by developing a computational method to compare brain organization across species.
Methods
The core of this study is a cross-species surface registration framework. They use Multimodal Surface Matching (MSM), a sophisticated algorithm that aligns cortical surfaces based on features. Specifically, they leverage myelin maps (derived from T1w/T2w MRI ratios, which correlate with cortical myelin content) as the primary feature for registration. This allows them to create a "common space" where homologous brain areas across macaque, chimpanzee, and human are aligned.
The key innovation is then using this myelin-based registration to predict the cortical projection patterns of white matter tracts. They compare these myelin-predicted tract maps with actual tract maps derived from diffusion MRI tractography. The degree of overlap (or lack thereof) between the predicted and actual tract maps reveals whether evolutionary changes in brain area location (captured by myelin registration) are sufficient to explain tract connectivity differences, or if the tracts themselves have undergone independent evolutionary modifications (e.g., extending into new cortical territories).
Datasets
The study utilizes neuroimaging data from three primate species:
- Human: 20 healthy volunteers (3T Siemens Prisma scanner).
- Chimpanzee: In vivo structural MRI and DW-MRI data from the National Chimpanzee Brain Resource (Yerkes National Primate Research Center, Emory University). Data from 29 adult chimpanzees were used for myelin maps.
- Macaque: Ex vivo DW-MRI data from four rhesus macaques and in vivo T1w/T2w scans from five macaques.
The paper also makes available processing code and group-level myelin maps and tract surface maps for all three species.
Key Insights and Ideas
- Myelin as a powerful anchor for cross-species alignment: Myelin maps provide reliable landmarks across primate species, enabling robust registration of homologous cortical areas.
- Dissociating cortical relocation from tract extension: The framework effectively separates evolutionary changes in brain area position from changes in white matter tract trajectories.
- Most tracts follow cortical relocation: For several major white matter tracts (MDLF, SLF3, IFO, and to some extent ILF), evolutionary changes in cortical area size and position adequately predict their connectivity patterns across species.
- The Arcuate Fasciculus is special: The arcuate fasciculus (AF), a key tract in language processing, stands out. Its connectivity pattern in humans cannot be fully explained by cortical relocation alone, suggesting it underwent independent evolutionary modifications.
- A flexible framework for evolutionary neuroscience: The approach is generalizable to other brain features and species, offering a powerful tool for testing comparative hypotheses about brain evolution.
Key Results
- The myelin-based registration successfully aligned homologous cortical areas across macaque, chimpanzee, and human, capturing significant distortions related to areal expansion.
- For tracts like the Middle Longitudinal Fasciculus (MDLF), Superior Longitudinal Fasciculus 3 (SLF3), and Inferior Fronto-Occipital Fasciculus (IFO), the myelin-based registration predicted their posterior terminations in the temporo-parietal cortex with high accuracy (high Dice coefficients, tract extension ratios close to 1). This indicates that cortical relocation was the primary driver of evolutionary change for these tracts.
- The Arcuate Fasciculus (AF) showed the lowest Dice coefficient and the highest tract extension ratio, particularly in macaque-to-human comparisons. This strongly suggests that the AF's connectivity pattern in humans evolved beyond simple relocation of cortical areas, indicating independent tract modifications.
- Minor extensions were also noted for the Inferior Longitudinal Fasciculus (ILF), suggesting some tract-specific evolution.
Limitations and Future Work
- Resolution limitations: The authors note that their method is limited by the spatial resolution of brain areas and cannot assign homology at finer scales.
- Frontal cortex registration: Myelin contrast is less pronounced in the human frontal cortex, making registration in these areas more challenging. Future work could incorporate other modalities (e.g., neurite orientation dispersion and density imaging - NODDI) for better frontal lobe analysis.
- Broader species and modalities: The authors envision expanding this framework to a wider range of species and incorporating additional neuroimaging modalities to gain a more comprehensive understanding of primate brain evolution.
- Investigating other tracts: While the study focused on specific tracts, future work could explore a broader set of white matter pathways.
- Functional implications: While the study highlights structural differences, future research could link these findings to functional specializations and behavioral differences across species.
Summary updated