Submission

Our project quantifies the integration-segregation balance in brain networks across the lifespan. We introduce metrics—Integration-Segregation Difference and Sum—to measure balance at network and regional levels. Using public datasets, we investigate how integration and segregation change during development, adulthood, and aging. Additionally, we identify which brain networks primarily drive these changes, and whether specific thalamic nuclei exhibit distinct patterns. We also explore how this balance correlates with cognitive and emotional capacities, personality traits, and mental health, along with its relationship to neurotransmitter levels and gene expression. These analyses offer insights into how brain network dynamics influence behavior, cognition, and mental well-being across the lifespan, providing a foundation for potential advances in addressing age-related neurological and mental health conditions.

Secondary Analysis Project

Our project aims to investigate how the integration-segregation balance in brain networks changes across the lifespan and its correlation with cognitive and emotional capacities, personality traits, and mental health. We recently proposed a novel network-based metric that measures the integration-segregation balance and validated it to anesthesia and sleep datasets (accepted in Nature Communications, to be published by Nov 2024). We will apply these metrics—Integration-Segregation Difference (ISD) and Sum (ISS)—to measure the dynamic balance between network integration and segregation at both the network and regional levels. This project will uncover how brain network dynamics influence cognition and mental well-being, offering insights into age-related neurological and mental health conditions. Additionally, we will explore relationships between network balance and molecular factors such as neurotransmitter levels and gene expression.

Data Utilization

We will utilize functional connectivity datasets (young adults and aging) from the Zenodo repository (https://zenodo.org/records/6770120) and an fMRI dataset covering development (5-21 years) from the Human Connectome Project. We will also incorporate neurotransmitter data and gene expression data from the Allen Brain Atlas, providing molecular insights relevant to our analyses.

Incorporation of GREI Repository Data

We plan to use the functional connectivity dataset for resting-state fMRI (young adults and aging) available in the Zenodo repository, which participates in the GREI.

Methods and Analysis

We will apply network neuroscience methods to analyze brain connectivity patterns. Using ISD and ISS metrics, we will assess how the balance between integration and segregation changes over development and aging in large-scale brain networks and specific regions. Based on the functional connectivity matrices from the Zenodo repository, we will calculate network- and region-level metrics, quantifying differences and sum in integration and segregation.

Additionally, we will perform correlation analyses between network balance metrics and sex, cognitive, emotional, and personality traits from the HCP datasets. We will explore relationships with neurotransmitter levels and gene expression using dominance analysis and gene ontology analysis.

Suggested Timeline

1. Month 1: Data acquisition and preprocessing, including downloading from HCP and GREI repositories.
2. Month 2: Calculation of network-level integration-segregation balance and benchmarking against existing metrics.
3. Month 3: Region-level analysis and initial result review.
4. Month 4-5: Correlation analyses with cognitive, emotional, and personality measures, as well as molecular data. Drafting publications and preparing conference presentations.
5. Months 6: Final review, validation, and additional analyses for replicability and reproducibility. Manuscript drafting for peer-reviewed journals.

Research Findings and Expected Outcomes

Our project aims to quantify how the integration-segregation balance in brain networks changes across the lifespan. By analyzing public lifespan datasets, we expect to uncover patterns of change in brain network dynamics during development, adulthood, and aging. Specifically, we anticipate that integration will decrease during development and more gradually in aging, while segregation may show a rapid decline in later stages of life. We expect these changes to be driven primarily by subcortical networks, with distinct patterns emerging in the unimodal and transmodal nuclei within the thalamus. Additionally, we anticipate that deviations in network balance will correlate with alterations in cognitive function, emotional regulation, and psychological well-being across the lifespan. Moreover, we expect to find links between these brain network dynamics and molecular factors, such as neurotransmitter levels and gene expression.

Sharing of the Findings

We are committed to sharing our findings with the scientific community. Key dissemination avenues include:

1. Peer-reviewed Publications: We plan to publish in high-impact journals related to neuroscience and cognitive science, detailing our novel metrics, analysis of lifespan datasets, and the correlation of network balance with cognitive and emotional outcomes.
2. Conference Presentations: We will present our findings at major conferences such as the Society for Neuroscience (SfN) and the Organization for Human Brain Mapping (OHBM). 

FAIR Principles

We fully adhere to the FAIR principles, ensuring our data, code, and results are openly available and reusable. Key steps include:

1. Findable: We will assign DOIs to datasets and code shared on platforms like OpenNeuro and GitHub.
2. Accessible: All data and code will be stored in public repositories without restrictions, using accepted formats for easy accessibility.
3. Interoperable: We will ensure our data and code follow community standards and provide clear metadata and annotations to ensure compatibility with other research.
4. Reusable: Our methodologies will be well-documented, with clear instructions for reuse. We will adopt open licensing to encourage widespread use.

Replicability and Reproducibility

We prioritize replicability and reproducibility through these actions:

1. Open Code and Data: Sharing all datasets and analysis pipelines will allow other researchers to replicate our analyses. Detailed documentation will ensure step-by-step reproducibility.
2. Version Control: We will use GitHub to track all versions of our code, enabling others to replicate our work and make improvements or adaptations as necessary.
3. Clear Methodological Reporting: Our publications will include thorough methodological details, covering data preprocessing and statistical analyses. We will follow best practices for reporting to ensure transparency and reproducibility.

Our project aims to quantify the integration-segregation balance in brain networks across the lifespan, contributing to network neuroscience, cognitive neuroscience, and neurobiology. By introducing novel network metrics (ISD and ISS), we offer new insights into how brain network dynamics evolve across development, adulthood, and aging. These metrics allow us to assess how different brain regions and networks interact to support cognitive and emotional functions and relate to molecular factors like neurotransmitters and gene expression.

Contributions to Scientific Disciplines

1. Network Neuroscience: We contribute to network neuroscience by developing tools that capture the dynamic balance of brain networks. There is an ongoing debate in current literature regarding changes in brain network integration and segregation during aging. Our novel ISD and ISS metrics help resolve this conflict by providing a comprehensive measurement of both processes. Therefore, our approach offers a more balanced and nuanced understanding of how brain networks interact and reorganize as individuals age. 
2. Cognitive neuroscience: By correlating brain network balance with cognitive and emotional capacities, we provide insights into how changes in network architecture impact cognition, memory, attention, and emotional regulation. 
3. Molecular neurobiology: We also explore how regional contributions to global network balance relate to neurobiological factors, including neurotransmitter levels and gene expression. This interdisciplinary approach bridges the gap between brain connectivity and molecular neurobiology.

Impact on Diagnosis, Treatment, and Prevention

1. Tracking Cognitive Decline: Our metrics offer insights into changes in brain network dynamics that may help track cognitive health over time. By identifying deviations in integration and segregation, we may detect early signs of cognitive impairments during development or aging, potentially serving as indicators of neural dysfunction.
2. Targeted Treatment: Understanding how brain networks reorganize during aging and disease provides insights that could inform treatment approaches. Recognizing regions with imbalanced integration or segregation might lead to strategies for neuromodulation aimed at restoring balance, potentially improving outcomes for cognitive and emotional impairments.
3. Mental Health Insights: Our findings may offer insights into mental health conditions, such as anxiety disorder and depression, where abnormal brain network balance is a factor. Identifying these imbalances may guide intervention strategies, such as cognitive training or lifestyle modifications

By advancing the understanding of brain network dynamics and their relationship with cognitive and biological factors, we lay the groundwork for future innovations in preserving cognitive function and improving mental health throughout the lifespan.

Our team consists of two members: Zirui Huang, Ph.D., and Hyunwoo Jang, M.S., both based at the University of Michigan Center for Consciousness Science. We have been collaborating closely for the past three years. Zirui is a Research Assistant Professor in Anesthesiology with over 50 publications and extensive experience in analyzing fMRI data, particularly in the context of consciousness and anesthesia. Hyunwoo, a Ph.D. student in the University of Michigan Neuroscience Graduate Program, is advised by Zirui and specializes in network neuroscience, focusing on the integration-segregation balance of brain networks.

Our collaboration is highly integrated, as we work in the same building, meeting daily to discuss progress and research goals. Our most recent joint work, accepted in Nature Communications (to be published by Nov 2024), examined integration-segregation balance changes during anesthesia and sleep, further solidifying our expertise in this area. Together, we bring a comprehensive skill set in statistical analysis, neuroimaging, and network-based approaches to tackle complex brain dynamics. Our strong teamwork, underpinned by close communication and shared expertise, positions us well to successfully complete the proposed project.

1. Data Quality and Selection: We use high-quality public lifespan datasets (HCP-Development, Young Adults, and Aging), as well as neurotransmitter and transcriptome datasets, providing robust coverage across life stages for comprehensive analysis.

2. Methodological Rigor: We employ novel network metrics—Integration-Segregation Difference (ISD) and Sum (ISS)— to quantify balance at network and regional levels. Our pipeline ensures methodological rigor, benchmarking the metrics’ performance using established methods.

3. Collaborative Expertise: Our team’s expertise in fMRI data analysis, network neuroscience, and statistical modeling ensures we can handle the project’s complexity. Daily interactions drive a dynamic, productive collaboration.

4. Ethics and Data Integrity: We adhere to ethical standards, ensuring non-discriminatory data use and maintaining confidentiality. We are committed to open science, sharing results and methods to ensure transparency and reproducibility.