Towards in vivo histology using quantitative MRI: application and validation strategies
Today, morphometric changes (i.e. volume loss) measured with non-invasive MRI using standard structural imaging methods (e.g. T1-weighted MR images) are well-established biomarkers for cognitive decline due to neurological disorders and normal aging. However, there is a need for a better understanding of the neurobiological mechanisms associated with these extensive morphometric changes. In vivo histology based on quantitative MRI (hMRI) can help to establish the missing link between measured MRI signals and the underlying tissue microstructure. This will be done by developing advanced biophysical models to relate the MR signal to underlying tissue properties such as fibre density, myelin density, or the g-ratio of fibre pathways (i.e. the ratio of the inner and outer diameter of myelinated axons) which are all key factor for neuronal conduction velocities and thus of functional relevance. Here, the idea behind hMRI is introduced and illustrated in specific applications. Moreover, validation strategies that combine gold standard ex vivo histology and deep learning are discussed.