Characterizing dynamic functional connectivity in congenitally blind people using Hidden Markov Models

Caterina Annalaura Pedersini,Bas Rokers,Olivier Collignon

DOI: https://doi.org/10.1101/2024.10.11.617823

2024-10-12

Abstract:How does sensory experience influence the brain connectome? Early acquired blindness is known to trigger a large-scale alteration in brain connectivity. Previous studies have traditionally assessed stationary functional connectivity across all timepoints in scanning sessions. In this study we compared the dynamic nature of functional connectivity in sighted (50) and early/congenitally blind (33) individuals through the data-driven approach of Hidden Markov Models (HMM). Blind individuals showed increased fractional occupancy in two key brain states: one characterized by bilateral somatomotor activation and another by occipito-frontal activation, mainly in the left hemisphere. Conversely, sighted individuals more frequently occupied a state associated with Default Mode and Ventral Attentional networks. Network analysis revealed decreased connectivity in blind individuals within states frequently occupied by sighted individuals, particularly between visual and somatomotor regions. In contrast, states more often visited by blind individuals exhibited increased connectivity within the visual network, connecting multiple occipital regions with the temporal fusiform area, as well as between visual and prefrontal regions. These results suggest that the frequency of visiting specific brain states may contribute to the differences observed between blind and sighted individuals in stationary resting state functional connectivity. Together, these findings show the effectiveness of Hidden Markov Models in capturing the temporal dynamics of functional connectivity, revealing its variability over time. We found that early visual deprivation disrupts these dynamics, with blindness leading to prolonged activation of unimodal regions and increased connectivity within the visual network and between occipital and higher-order brain regions. Additionally, alterations in the chronnectome, defined as the functional connectome changing over time - specifically in the time spent visiting a state - may provide new insights into stationary functional connectivity patterns in blind individuals, establishing a foundation for understanding network connectivity dynamics following sensory deprivation.

Neuroscience

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to explore how the dynamic characteristics of functional connectivity in congenitally blind people are different from those in the normal - sighted population, especially by using Hidden Markov Models (HMM) to capture these differences. Specifically, the researchers hope to evaluate whether early visual deprivation leads to changes in the temporal dynamics of functional connectivity by analyzing resting - state functional magnetic resonance imaging (resting - state fMRI) data, and explore whether these dynamic changes can provide new insights into the differences in static functional connectivity patterns, especially in terms of the reduced connectivity between the visual cortex and the sensorimotor areas.

The specific objectives of the paper include:
1. **Evaluate the impact of early visual deprivation on the temporal dynamics of functional connectivity**: The researchers hypothesize that early visual deprivation will trigger changes in the temporal dynamics of specific brain regions (especially unimodal regions).
2. **Explore the explanatory power of dynamic changes for differences in static functional connectivity patterns**: The researchers aim to evaluate whether these dynamic changes can explain the abnormalities in the static functional connectivity patterns observed in congenitally blind people, especially the reduction in connectivity between the visual cortex and the sensorimotor areas.

To achieve these goals, the researchers adopted the HMM method to analyze the resting - state fMRI data from congenitally blind and normal - sighted individuals. By comparing indicators such as the residence time and transition probability in different brain states between the two groups of people, they revealed that congenitally blind people have an increased residence time in some key brain states, as well as patterns of enhanced or weakened connectivity between brain regions in these states. These findings not only demonstrate the effectiveness of HMM in capturing the temporal dynamics of functional connectivity, but also provide a new perspective for understanding the changes in brain network connectivity dynamics after sensory deprivation.