In this chapter, we have described several quantitative MRI measures, which are promising for the characterization of brain connectivity. However, to date, there has been a relative paucity of experiments that have directly compared functional and effective measures of brain connectivity with these structural and anatomical measures of brain connectivity and physiology. This is likely to change in the near future as these techniques become more available. Characterization of WM anatomy and physiology with MRI may enable more complex models of brain connectivity to be developed, as the circuitry of brain connectivity becomes more well-defined. For example, many have proposed that FA increases are primarily reflecting myelination. This leads to the prediction that FA would be correlated most strongly with the short myelin-water fraction from T2 relaxometry experiments, as well as to the size of the macro-molecular pool in quantitative MT studies. On the other hand, if changes in fibre density underlie changes in FA (Beaulieu 2002), FA should be more strongly associated with the extracellular water peak. Preliminary evidence for this prediction comes from recent work showing FA was not correlated with the myelin water fraction in white matter (MacKay et al. 2006). Functional and effective connectivity studies so far have generally modeled the brain as a "black box" with inputs and outputs, and most of the internal circuitry has been derived from non-human primate studies. Quantitative structural and physiological image data from MRI may provide critical information about the functional circuitry within the black box.
To move quantitative MRI into the forefront of techniques for characterizing brain connectivity, further developments are necessary. Obviously, improvements to both the imaging technology through better and more efficient pulse sequences, imaging RF coils, and gradient coils, and quantitative imaging models and image analysis methods will facilitate comparisons between more conventional connectivity measures with quantitative MRI measures of WM. However, even with improvements in the technology, the application will be somewhat limited unless they become more readily available, either through the MRI system manufacturers or through research collaborations. While the methodologies are still young and emerging, we can already pose some interesting questions: Do variations in diffusion parameters or myelin content along tracts relate to function? Is the whole fibre tract affected? Does knowing something about tract likelihood help predict differences in functional and effective connectivity? Answers to these and similar questions will require multimodal imaging, as most quantitative MRI studies have focused on a single measure or measurement type We will also need a better understanding of the statistical properties of the data, and sophisticated multivariate and nonlinear modeling techniques, some of which are already available, and others of which are discussed throughout this volume. This will be an iterative process and will require refinement of both imaging and analysis techniques. However, we have optimism that in the end the model fits will be acceptable and we will know something useful about how brain structure contributes to brain function.
Was this article helpful?