Gemma Nedjati-Gilani

A ranking of diffusion MRI compartment models with in vivo human brain data.

Diffusion magnetic resonance imaging (MRI) microstructure imaging provides a unique noninvasive probe into tissue microstructure. The technique relies on biophysically motivated mathematical models, relating microscopic tissue features to the magnetic resonance (MR) signal. This work aims to determine which compartment models of diffusion MRI are best at describing measurements from in vivo human brain white matter.

Machine learning based compartment models with permeability for white matter microstructure imaging

Abstract Some microstructure parameters, such as permeability, remain elusive because mathematical models that express their relationship to the MR signal accurately are intractable. Here, we propose to use computational models learned from simulations to estimate these parameters. We demonstrate the approach in an example which estimates water residence time in brain white matter. The residence time &tgr;i of water inside axons is a potentially important biomarker for white matter pathologies of the human central nervous system, as myelin damage is hypothesised to affect axonal permeability, and thus &tgr;i. We construct a computational model using Monte Carlo simulations and machine learning (specifically here a random forest regressor) in order to learn a mapping between features derived from diffusion weighted MR signals and ground truth microstructure parameters, including &tgr;i. We test our numerical model using simulated and in vivo human brain data. Simulation results show that estimated parameters have strong correlations with the ground truth parameters (Symbol) for volume fraction, residence time, axon radius and diffusivity respectively), and provide a marked improvement over the most widely used Kärger model (Symbol). The trained model also estimates sensible microstructure parameters from in vivo human brain data acquired from healthy controls, matching values found in literature, and provides better reproducibility than the Kärger model on both the voxel and ROI level. Finally, we acquire data from two Multiple Sclerosis (MS) patients and compare to the values in healthy subjects. We find that in the splenium of corpus callosum (CC‐S) the estimate of the residence time is 0.57±0.05 s for the healthy subjects, while in the MS patient with a lesion in CC‐S it is 0.33±0.12 s in the normal appearing white matter (NAWM) and 0.19±0.11 s in the lesion. In the corticospinal tracts (CST) the estimate of the residence time is 0.52±0.09 s for the healthy subjects, while in the MS patient with a lesion in CST it is 0.56±0.05 s in the NAWM and 0.13±0.09 s in the lesion. These results agree with our expectations that the residence time in lesions would be lower than in NAWM because the loss of myelin should increase permeability. Overall, we find parameter estimates in the two MS patients consistent with expectations from the pathology of MS lesions demonstrating the clinical potential of this new technique. Symbol. No caption available. Symbol. No caption available. HighlightsSome tissue parameters remain elusive because mathematical models are intractable.We propose to use machine learning to estimate these parameters, here permeability.Simulation results show an excellent agreement between estimations and ground truth.New technique performs better than the standard Karger Model.In‐vivo results consistent with pathology of MS lesions showing clinical potential.

Machine Learning Based Compartment Models with Permeability for White Matter Microstructure Imaging

The residence time Ti of water inside axons is an important biomarker for white matter pathologies of the human central nervous system, as myelin damage is hypothesised to increase axonal permeability, and thus reduce Ti. Diffusion-weighted (DW) MRI is potentially able to measure Ti as it is sensitive to the average displacement of water molecules in tissue. However, previous work addressing this has been hampered by a lack of both sensitive data and accurate mathematical models. We address the latter problem by constructing a computational model using Monte Carlo simulations and machine learning in order to learn a mapping between features derived from DW MR signals and ground truth microstructure parameters. We test our method using simulated and in vivo human brain data. Simulation results show that our approach provides a marked improvement over the most widely used mathematical model. The trained model also predicts sensible microstructure parameters from in vivo human brain data, matching values of Ti found in the literature.

Tissue Microstructure Imaging with Diffusion MRI

Over the past 20 years, diffusion magnetic resonance imaging has been established as one of the key methods for noninvasive imaging of tissue microstructure. This article provides an overview of the current state-of-the-art techniques for estimating features such as fiber orientation, dispersion, size, and water exchange rate and provides some perspectives on the future of the field.

Ranking diffusion-MRI models with in-vivo human brain data

Diffusion MRI microstructure imaging provides a unique non-invasive probe into the microstructure of biological tissue. Its analysis relies on mathematical models relating microscopic tissue features to the MR signal. This work aims to determine which compartment models of diffusion MRI are best at describing the signal from in-vivo brain white matter. Recent work shows that three compartment models, including restricted intra-axonal, glial compartments and hindered extra-cellular diffusion, explain best multi b-value data sets from fixed rat brain tissue. Here, we perform a similar experiment using in-vivo human data. We compare one, two and three compartment models, ranking them with standard model selection criteria. Results show that, as with fixed tissue, three compartment models explain the data best, although simpler models emerge for the in-vivo data. We also find that splitting the scanning into shorter sessions has little effect on the models fitting and that the results are reproducible. The full ranking assists the choice of model and imaging protocol for future microstructure imaging applications in the brain.