Research
Print page Print page
Switch language
The Capital Region of Denmark - a part of Copenhagen University Hospital
Published

Predicting the fMRI Signal Fluctuation with Recurrent Neural Networks Trained on Vascular Network Dynamics

Research output: Contribution to journalJournal articleResearchpeer-review

DOI

  1. Multimodal Image Analysis of Apparent Brain Age Identifies Physical Fitness as Predictor of Brain Maintenance

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. Is There a Correlation Between the Number of Brain Cells and IQ?

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Human Cerebral Perfusion, Oxygen Consumption, and Lactate Production in Response to Hypoxic Exposure

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. Dissociable and Paradoxical Roles of Rat Medial and Lateral Orbitofrontal Cortex in Visual Serial Reversal Learning

    Research output: Contribution to journalJournal articleResearchpeer-review

  5. Uncovering a Role for the Dorsal Hippocampal Commissure in Recognition Memory

    Research output: Contribution to journalJournal articleResearchpeer-review

  1. In vivo tensor-valued diffusion MRI of focal demyelination in white and deep grey matter of rodents

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. High spatiotemporal vessel-specific hemodynamic mapping with multi-echo single-vessel fMRI

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Streamline tractography for 3D mapping of axon bundle organization in one MRI voxel using ultra-high resolution synchrotron radiation imaging

    Research output: Chapter in Book/Report/Conference proceedingConference abstract in proceedingsResearchpeer-review

  4. Multidimensional Diffusion MRI Assists Myelin-sensitive Bound Pool Fraction in Differentiating Microstructural Maturity of Primate Brains

    Research output: Chapter in Book/Report/Conference proceedingConference abstract in proceedingsResearchpeer-review

  5. Potential diagnostic value of a type X collagen neo-epitope biomarker for knee osteoarthritis

    Research output: Contribution to journalJournal articleResearchpeer-review

  • Filip Sobczak
  • Yi He
  • Terrence J Sejnowski
  • Xin Yu
View graph of relations

Resting-state functional MRI (rs-fMRI) studies have revealed specific low-frequency hemodynamic signal fluctuations (<0.1 Hz) in the brain, which could be related to neuronal oscillations through the neurovascular coupling mechanism. Given the vascular origin of the fMRI signal, it remains challenging to separate the neural correlates of global rs-fMRI signal fluctuations from other confounding sources. However, the slow-oscillation detected from individual vessels by single-vessel fMRI presents strong correlation to neural oscillations. Here, we use recurrent neural networks (RNNs) to predict the future temporal evolution of the rs-fMRI slow oscillation from both rodent and human brains. The RNNs trained with vessel-specific rs-fMRI signals encode the unique brain oscillatory dynamic feature, presenting more effective prediction than the conventional autoregressive model. This RNN-based predictive modeling of rs-fMRI datasets from the Human Connectome Project (HCP) reveals brain state-specific characteristics, demonstrating an inverse relationship between the global rs-fMRI signal fluctuation with the internal default-mode network (DMN) correlation. The RNN prediction method presents a unique data-driven encoding scheme to specify potential brain state differences based on the global fMRI signal fluctuation, but not solely dependent on the global variance.

Original languageEnglish
JournalCerebral Cortex
Volume31
Issue number2
Pages (from-to)826-844
Number of pages19
ISSN1047-3211
DOIs
Publication statusPublished - 1 Feb 2021

    Research areas

  • default mode network, machine learning, resting state, single-vessel, slow oscillation

ID: 60880916