Related Experiment Video
Updated: Jul 11, 2025
09:59
A Magnetic Resonance Imaging Protocol for Stroke Onset Time Estimation in Permanent Cerebral Ischemia
Published on: September 16, 2017
14.1K
Current status in spatiotemporal analysis of contrast-based perfusion MRI
Eve S Shalom1,2, Amirul Khan3, Sven Van Loo1,4
1School of Physics and Astronomy, University of Leeds, Leeds, UK.
Magnetic Resonance in Medicine
|November 6, 2023
Summary
Perfusion MRI analysis is shifting from isolated voxels to interconnected systems. Machine learning now makes complex spatiotemporal modeling feasible, improving accuracy in perfusion MRI studies.
Area of Science:
- Medical Imaging
- Biophysics
- Computational Neuroscience
Background:
- Current perfusion MRI analysis treats voxels as isolated, leading to systematic errors.
- This ignores the spatially organized network and indicator exchange between neighboring voxels.
- Existing models for inter-voxel interactions are computationally intractable.
Purpose of the Study:
- To review spatiotemporal modeling approaches for perfusion MRI.
- To establish a coherent nomenclature and notation for these advanced methods.
- To clarify the state-of-the-art and identify future research priorities.
Main Methods:
- Review of existing literature on spatiotemporal modeling in perfusion MRI.
- Harmonization of nomenclature and notation for inter-voxel exchange models.
- Discussion of machine learning advancements enabling complex inverse problems.
Main Results:
- Identified limitations of the isolated-voxel paradigm in perfusion MRI.
- Highlighted the potential of spatiotemporal models for improved accuracy.
- Demonstrated the feasibility of previously intractable inverse problems with machine learning.
Conclusions:
- A paradigm shift towards network-based perfusion MRI analysis is emerging.
- Machine learning is crucial for implementing advanced spatiotemporal models.
- Further research is needed to refine these methods and address knowledge gaps.
