ABSTRACT
Neuroimaging studies aim to identify markers that capture inter-individual differences in behaviour, function, dysfunction or pathology. To date, most computational neuroanatomy approaches have utilised conventional weighted imaging to achieve this, for example, analysing morphometry based on T1-weighted images. However, weighted images depend on many physical tissue properties (e.g. relaxation rates, susceptibility and proton density) as well as instrumental modulations (e.g. receive field inhomogeneities) making the interpretation of changes based on such images complex. Quantitative MRI (qMRI) data provide more specific measures by mapping an individual tissue property, for example, the longitudinal relaxation rate – corrected for instrumental bias. This chapter discusses how established and emerging concepts in qMRI can help improve our understanding of identified differences, beginning by revisiting the utility and pitfalls of qMRI in neuroscientific applications. Common artefacts, such as participant motion, low signal-to-noise ratio and susceptibility-based distortions, and correction schemes are discussed. In the second part of this chapter, methods for analysing qMRI data at the group or inter-individual level are discussed. Particular care must be taken when nonlinear warping procedures are applied to analyse these data in a template space, for example, MNI space. Therefore, some of the methods that aim to preserve quantitative metrics during such manipulation are highlighted.
Chapter 18: The future of quantitative MRI
