Mapping directed information flow between homotopic regions of the human brain

Introduction: Homotopic connectivity is an integral component of the brain’s functional architecture, typically quantified using the Pearson correlation between mirrored voxels or parcels [1]. Despite mounting evidence for left-right structural asymmetry in health [2,3] and disease [4,5], interhemispheric information flow is seldom directly assessed.

Methods: We quantified homotopic directed information (DI) from functional magnetic resonance imaging (fMRI) to measure directed interhemispheric coupling for the first time [6]. All resting-state fMRI data were obtained the Human Connectome Project (N=100) and were preprocessed and parcellated into the Desikan-Killiany cortical atlas in [7]. For each of 34 cortical regions (one per hemisphere), we computed DI with a Gaussian density estimator using pyspi [8] in Python.

Results: Homotopic DI magnitudes varied across the brain, ranging from the lowest average magnitude in the entorhinal cortex (1.23 × 10-2 ± 8.59 × 10-3) to the highest in the lateral occipital cortex (2.39 ± 0.96). Generally, medial occipital and superior parietal regions exhibited higher homotopic DI while inferior frontal and temporal regions exhibited lower homotopic DI. We also compared left-right asymmetries, finding that lateral regions exhibit greater DI from left to right while medial regions exhibit greater DI from right to left.

Conclusions: Our findings suggest that there are gradients of both the magnitude and direction of interhemispheric information flow, warranting further characterization of the functional implications for health and disease.

References: [1] Mancuso et al. 2019. Scientific Reports; [2] Wan et al. 2024. bioRxiv; [3] Kong, et al. 2022. Human Brain Mapping; [4] Chen et al. 2024. Brain Communications; [5] Roe et al. 2021. Nature Communications; [6] Lizier et al. 2011, Journal of Computational Neuroscience; [7] Fallon et al. 2020. Network Neuroscience; [8] Cliff et al. 2023. Nature Computational Science