This research article introduces a novel method for mapping the human brain by examining metabolic similarity matrices (MetSiMs) to understand how different regions share chemical profiles. The authors utilized advanced MRSI imaging technology to track various metabolites across two independent participant groups from Geneva and Lausanne. By applying dimensionality reduction techniques like t-SNE, the study identifies specific metabolic modes that characterize local and global brain patterns. Interestingly, the data reveals only a modest correlation between these chemical similarities and the brain's physical structural connections. Ultimately, the work provides a sophisticated mathematical framework for exploring how the brain's biochemical landscape organizes itself independently of traditional anatomical pathways.

References:

• Lucchetti F, Céléreau E, Steullet P, et al. Constructing the human brain metabolic connectome with MR spectroscopic imaging reveals cerebral biochemical organization[J]. Nature Communications, 2025.