Conserved and variable architecture of human white matter connectivity

Area of Science:

• Neuroscience
• Medical Imaging
• Network Science

Background:

• Whole-brain network analysis of diffusion imaging tractography quantifies brain connectivity patterns.
• Understanding network architecture reproducibility is crucial for reliable individual and group comparisons.

Purpose of the Study:

• To investigate the conservation of network properties across methodological variations.
• To assess the reproducibility of individual brain architecture across multiple scanning sessions.

Main Methods:

• Diffusion spectrum imaging (DSI) and diffusion tensor imaging (DTI) data acquired in triplicate from healthy adults.
• Deterministic tractography performed on each dataset.
• Inter-regional connectivity matrices derived using three parcellation schemes across spatial resolutions.

Main Results:

• Anatomical brain networks exhibit sparsity, hierarchy, and assortativity, with signatures of topo-physical interdependence (e.g., Rentian scaling).
• Connectivity properties and graph metrics showed high reproducibility and low variability for both DSI and DTI.
• DSI demonstrated higher sensitivity to complex fiber configurations, resulting in increased tract counts and network density compared to DTI.

Conclusions:

• Network analysis of human white matter connectivity offers sensitive and temporally stable topological and physical estimates of individual cortical structure.
• Results confirm the robustness of network analysis across different diffusion imaging techniques and parcellation strategies.
• The findings support the utility of diffusion imaging-based network analysis for characterizing individual brain organization.