Purpose: Multi-echo gradient recalled echo MRI (mGRE-MRI) has been shown to contain information on signal compartments in white matter and, compartments have been associated with myelin, axonal and extracellular water. We recently showed that signal compartments can be present in different tissue types in the human brain. It is however unclear what mGRE-MRI signal compartmentalisation means in general within the brain and how it can be used for diagnosing and monitoring neurodegenerative diseases and disorders.
Methods: We used mGRE-MRI data in a cohort of healthy participants to investigate whether the unexpected trends in magnetic susceptibility, mapped using quantitative susceptibility mapping, can be explained through signal compartmentalisation. We also implemented a data driven approach to be able to define the number of signal compartments based on studying distinct brain regions, such as the caudate, putamen, corpus callosum, thalamus, substantia nigra, pallidum, fornix, insula and cerebrospinal fluid. Additionally, we investigated how signal compartments change in focal epilepsy, in particular focal cortical dysplasia.
Results: We found mGRE-MRI signal compartmentalisation can explain the unexpected trend in quantitative susceptibility mapping as a function of echo time. Further analysis resulted in different signal compartments for each brain region, implying the presence of differential effects. These effects may lead to new biomarkers of neurodegenerative diseases and disorders. Using the data driven clustering approach, we identified five distinct mGRE-MRI signal compartments in brain regions investigated. This analysis led two key observations. Firstly, different brain regions can be defined using a different number of signal compartments. Secondly, signal compartments can be shared across different brain regions. By fixing signal compartment parameters we were able to spatially resolve the volume fraction of each compartment in the brain. We found a significant change in signal compartment parameters in the presence of focal cortical dysplasia.
Conclusions: Our work on mGRE-MRI signal compartmentalisation suggests the presence of new contrast mechanisms generated via the signal compartment parameters. These parameters, when resolved spatially across the human brain, may lead to new biomarkers of diseases and disorders affecting the central nervous system. Using the example of focal epilepsy, we were able to show systematic changes in mGRE-MRI compartment parameters, which may become useful in delineating brain regions affected by focal cortical dysplasia.