This study investigates the relationship between the shape of the brain's white matter connections and individual language performance. The key highlights are:
The study uses diffusion MRI tractography to reconstruct the brain's white matter connections as sequences of 3D points, which are then grouped into fiber clusters with different anatomical shapes.
In addition to traditional tissue microstructure and connectivity features, the study extracts 12 shape descriptors for each fiber cluster, including length, diameter, elongation, volume, surface area, and irregularity.
The authors introduce a novel deep learning framework called SFFormer that leverages a multi-head cross-attention module to fuse features from the shape, microstructure, and connectivity domains.
Experiments on a large dataset of 1065 healthy young adults show that the shape features, both individually and when fused with other features, outperform traditional microstructure and connectivity features in predicting individual language performance, as measured by vocabulary comprehension and oral reading tests.
The results indicate that the shape of the brain's white matter connections is an important factor in understanding and predicting human language function.
To Another Language
from source content
arxiv.org
Key Insights Distilled From
by Yui Lo,Yuqia... at arxiv.org 03-29-2024
https://arxiv.org/pdf/2403.19001.pdfDeeper Inquiries