In clinical practice, the interpretation of conventional MR images largely relies on the visual assessment of contrast or enhancement pattern of anatomical structures, where an expert bases the assessment on the basis of these differences. This contrast depends on the choice of relevant imaging parameters such as the repetition or echo times and the flip angle, but also on factors where one has less control such as vendor implemented post-processing, the B1+ field or receive field inhomogeneities. In quantitative MRI, by contrast, MR images are acquired and analyzed in a way that estimates one or more biological, chemical, or physical properties using a ratio scale. As one typically needs at least two images to provide this estimate, quantitative MRI is usually much slower compared to conventional MRI. On the other hand, the advantages of quantitative MRI include a reduced sensitivity to factors such as repetition or echo times, improved objectivity when interpreting data and an improved ability to compare findings within individuals over time and between different research or clinical sites. For quantitative MRI to come to full fruition, an estimation of the accuracy (closeness to the ground truth), precision (repeatability) and sensitivity (reproducibility) of the method are needed. In this talk, I will highlight a number of quantitative MRI techniques that are commonly used in skeletal muscle, along with their accuracy, sensitivity, reproducibility, and relation to a biological property. The most commonly used quantitative method to assess muscle volume and muscle contractile volume – ie the area of an individual muscle which is not replaced or infiltrated by fat, is based on the chemical-shift difference between water and fat. Originally introduced by Dixon [1], the technique has now matured and has been extended, and has been reviewed in detail elsewhere [2, 3]. This same technique is often used to assess the fat-fraction of a skeletal muscle, which in neuromuscular disease is related to disease progression. Using MR spectroscopy, a differentiation can be made between intra-myocellular, which are metabolically active, and extra-myocellular lipids.
Information about muscle architecture is often obtained using diffusion based techniques. These techniques provide information on restrictions to diffusion and average diffusion in three dimensions via the ‘conventional’ parameters of fractional anisotropy and mean diffusivity, respectively. These metrics are sensitive to muscle characteristics such as fiber type and blood volume, and processes such as atrophy and inflammation, but they are not specific [4].
T2 relaxometry has been used extensively to obtain information on the presence of inflammation or edema, both in sports sciences as in neuromuscular disease, commonly by multi-echo-spin echo approaches or MR spectroscopy. An increase in the T2 relaxation time can be an indication of intra and/or extracellular edema or even ‘disease activity’ in the muscle tissue [5]. In muscle, the T2 relaxation of muscle water is around 30 ms at 3T, whereas the T2 of fat is much longer (>80 ms) and the increase in T2 relaxation time of inflamed/edematous tissue rarely exceeds 10 ms. Since in many muscle diseases, particularly in chronic stages, muscle tissue is progressively replaced by fat, assessment of the T2 relaxation time without accounting for the presence of fat will lead to an overestimation of the T2 of the muscle water and possible wrong conclusions about the presence of edema. Muscle is a very dynamic organ, and metabolic demand and perfusion can change up to 10-fold during exercise.
Dynamic measurements using in-magnet muscle contraction protocols during and after exercise can be performed with a multitude of techniques, including phosphorous MR spectroscopy, T2* or T2 weighted imaging or mapping, intra-voxel incoherent motion, arterial spin labelling and phase contrast MRI [5, 6].
Concluding remarks
Quantitative MR of muscles makes use of range of techniques, of which a subset was introduced here. As pathology is rarely limited to a single feature, there is a strong added value in combining different MR measurements, where assessments are co-localized based on pathology. This hopefully allows disentangling the different pathogenic processes.

Acknowledgements

No acknowledgement found.