Baker Cole, Suarez-Mendez Isabel, Smith Grace, Marsh Elisabeth B, Funke Michael, Mosher John C, Maestu Fernando, Xu Mengjia, Pantazis Dimitrios
IEEE J Biomed Health Inform. 2024 Dec;28(12):7357-7368. doi: 10.1109/JBHI.2024.3416890. Epub 2024 Dec 5.
An expansive area of research focuses on discerning patterns of alterations in functional brain networks from the early stages of Alzheimer's disease, even at the subjective cognitive decline (SCD) stage. Here, we developed a novel hyperbolic MEG brain network embedding framework for transforming high-dimensional complex MEG brain networks into lower-dimensional hyperbolic representations. Using this model, we computed hyperbolic embeddings of the MEG brain networks of two distinct participant groups: individuals with SCD and healthy controls. We demonstrated that these embeddings preserve both local and global geometric information, presenting reduced distortion compared to rival models, even when brain networks are mapped into low-dimensional spaces. In addition, our findings showed that the hyperbolic embeddings encompass unique SCD-related information that improves the discriminatory power above and beyond that of connectivity features alone. Notably, we introduced a unique metric-the radius of the node embeddings-which effectively proxies the hierarchical organization of the brain. Using this metric, we identified subtle hierarchy organizational differences between the two participant groups, suggesting increased hierarchy in the dorsal attention, frontoparietal, and ventral attention subnetworks among the SCD group. Last, we assessed the correlation between these hierarchical variations and cognitive assessment scores, revealing associations with diminished performance across multiple cognitive evaluations in the SCD group. Overall, this study presents the first evaluation of hyperbolic embeddings of MEG brain networks, offering novel insights into brain organization, cognitive decline, and potential diagnostic avenues of Alzheimer's disease.
一个广泛的研究领域聚焦于从阿尔茨海默病的早期阶段，甚至在主观认知下降（SCD）阶段，识别功能性脑网络的变化模式。在此，我们开发了一种新颖的双曲型脑磁图（MEG）脑网络嵌入框架，用于将高维复杂的MEG脑网络转换为低维双曲表示。使用该模型，我们计算了两个不同参与者组的MEG脑网络的双曲嵌入：SCD个体和健康对照。我们证明这些嵌入保留了局部和全局几何信息，与竞争模型相比呈现出减少的失真，即使当脑网络被映射到低维空间时也是如此。此外，我们的研究结果表明，双曲嵌入包含独特的与SCD相关的信息，其提高了辨别能力，超越了仅连接性特征的辨别能力。值得注意的是，我们引入了一个独特的度量——节点嵌入的半径——它有效地代表了大脑的层次组织。使用这个度量，我们识别了两个参与者组之间细微的层次组织差异，表明SCD组的背侧注意、额顶叶和腹侧注意子网络中的层次增加。最后，我们评估了这些层次变化与认知评估分数之间的相关性，揭示了与SCD组多项认知评估中表现下降的关联。总体而言，本研究首次对MEG脑网络的双曲嵌入进行了评估，为大脑组织、认知下降和阿尔茨海默病的潜在诊断途径提供了新的见解。
