Diffusion MRI microstructure models with in vivo human brain Connectome data: results from a multi‐group comparison

Abstract

A large number of mathematical models have been proposed to describe the measured signal in diffusion-weighted (DW) magnetic resonance imaging (MRI). However, model comparison to date focuses only on specific subclasses, e.g. compartment models or signal models, and little or no information is available in the literature on how performance varies among the different types of models. To address this deficiency, we organized the ‘White Matter Modeling Challenge’ during the International Symposium on Biomedical Imaging (ISBI) 2015 conference. This competition aimed to compare a range of different kinds of models in their ability to explain a large range of measurable in vivo DW human brain data. Specifically, we assessed the ability of models to predict the DW signal accurately for new diffusion gradients and b values. We did not evaluate the accuracy of estimated model parameters, as a ground truth is hard to obtain. We used the Connectome scanner at the Massachusetts General Hospital, using gradient strengths of up to 300 mT/m and a broad set of diffusion times. We focused on assessing the DW signal prediction in two regions: the genu in the corpus callosum, where the fibres are relatively straight and parallel, and the fornix, where the configuration of fibres is more complex. The challenge participants had access to three-quarters of the dataset and their models were ranked on their ability to predict the remaining unseen quarter of the data. The challenge provided a unique opportunity for a quantitative comparison of diverse methods from multiple groups worldwide. The comparison of the challenge entries reveals interesting trends that could potentially influence the next generation of diffusion-based quantitative MRI techniques. The first is that signal models do not necessarily outperform tissue models; in fact, of those tested, tissue models rank highest on average. The second is that assuming a non-Gaussian (rather than purely Gaussian) noise model provides little improvement in prediction of unseen data, although it is possible that this may still have a beneficial effect on estimated parameter values. The third is that preprocessing the training data, here by omitting signal outliers, and using signal-predicting strategies, such as bootstrapping or cross-validation, could benefit the model fitting. The analysis in this study provides a benchmark for other models and the data remain available to build up a more complete comparison in the future.

Microstructure Imaging of Crossing (MIX) White Matter Fibers from diffusion MRI

“…We hope that our proposed method might provide a new, more flexible computational framework to revisit and expand studies on biophysical models evaluation and classification1718, to improve current models4041, and to develop future contests such as the White Matter Modeling Challenge 42 (http://cmic.cs.ucl.ac.uk/wmmchallenge/). …”

Motion‐robust diffusion compartment imaging using simultaneous multi‐slice acquisition

“…In this study, we aimed to use this data to estimate a diffusion model that accounts for the intra‐voxel orientation heterogeneity and allows delineation of crossing white matter fascicles, believed to occur in 60 to 90% of brain white matter voxels . Without loss of generality, we formulate the problem based on the DIAMOND approach which performed well in a recent comparison of several state‐of‐the‐art DW‐MRI models . DIAMOND is a hybrid biophysical and statistical diffusion model that allows characterization of fascicle‐specific diffusion characteristics while also modeling the intra‐compartment heterogeneity.…”

“…In this paper, we extend a framework based on slice‐level motion correction by exploiting the rigid coupling between simultaneously acquired slices for robust reconstruction in DCI. Without loss of generality, we build our reconstruction framework based upon the DIAMOND model which performed well compared to other models in a recent multi‐group comparison of DCI models . Compared to the ball‐and‐stick model that we previously considered, DIAMOND provides a much better model of brain microstructure as it simultaneously estimates the number of compartments and fits a statistical distribution of diffusion tensors for each compartment, effectively allowing characterization of diffusivity rate for each fascicle while accounting for the microstructural heterogeneity.…”

“…restricted water diffusion compartments [13][14][15][16][17][18][19][20][21] in techniques that we refer to as Diffusion Compartment Imaging (DCI) 20 . Fitting more complex models in DCI requires relatively dense q-space sampling 18,22 which can be extremely time consuming to acquire.…”

“…Without loss of generality, we build our reconstruction framework based upon the DIAMOND model 20 which performed well compared to other models in a recent multi-group comparison of DCI models. 21 Compared to the ball-and-stick model 13 that we previously considered, 24 DIAMOND provides a much better model of brain microstructure as it simultaneously estimates the number of compartments and fits a statistical distribution of diffusion tensors for each compartment, effectively allowing characterization of diffusivity rate for each fascicle while accounting for the microstructural heterogeneity. In designing the motion correction component, we take advantage of the rigid coupling between simultaneously acquired slices in the SMS sequences routinely used for DCI for both intra-slice motion detection and slice-level motion tracking.…”