Related Experiment Video
Updated: Jan 3, 2026
Author Spotlight: Noninvasive Cerebral Blood Flow Determination in Human Functional Brain Region for Diagnosis of Neurological Disorders
Published on: May 31, 2024
Estimating regional cerebral blood flow using resting-state functional MRI via machine learning
Ganesh B Chand1, Mohamad Habes2, Sudipto Dolui1
1Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA; Department of Neurology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA.
This study introduces a machine learning method to estimate brain perfusion using resting-state fMRI (rsfMRI) signals. The findings suggest rsfMRI can provide valuable perfusion information, especially when advanced perfusion MRI is unavailable.
Area of Science:
- Neuroimaging
- Biomedical Engineering
- Machine Learning
Background:
- Perfusion MRI is crucial for studying brain aging and diseases but isn't always accessible.
- Cerebrovascular changes are key indicators in neurological conditions.
Purpose of the Study:
- To develop a novel method for estimating regional cerebral blood flow (CBF) using resting-state functional MRI (rsfMRI) spectral information.
- To validate this method against arterial spin labeling (ASL) in individuals with normal cognition and mild cognitive impairment (MCI).
Main Methods:
- Machine learning models (Support Vector Machines) were trained using paired rsfMRI and ASL data.
- Models estimated regional CBF from rsfMRI signals alone, focusing on specific frequency bands (0.01-0.15 Hz).
Main Results:
- The rsfMRI-based method showed significant associations between estimated and actual CBF in lobar regions (parietal, occipital) and 24 regions of interest.
- The superior parietal lobule demonstrated the highest correlation (r=0.50).
- Estimated CBF correlated with cognitive scores (MMSE) and was lower in MCI patients.
Conclusions:
- Resting-state fMRI signals contain quantifiable perfusion information.
- The proposed rsfMRI-based method can serve as a surrogate for perfusion imaging when ASL is not available.
- This approach may aid in assessing cerebrovascular health and cognitive decline.
