Currently​, only post mortem imaging methods, such as histology on slices, allow the characterization of brain cells morphology. Now, the morphology of neurons and astrocytes is susceptible to be altered during some pathologies, in particular neurodegenerative diseases. It would therefore be extremely relevant to have at our disposal a non-invasive imaging modality allowing access to these morphological characteristics. Unfortunately, the thinness of the structure of these cells, as well as their extreme entanglement, put them well below the resolution of all in vivo imaging techniques.

The group of Julien Valette has developed, in the framework of a project funded by the European Research Council (ERC), a strategy to quantify some of the morphological characteristics of neurons and astrocytes, as published in PNAS in May 2016. This indirect method is based on the measurement, by original in vivo magnetic resonance spectroscopy approaches, of the diffusion of endogenous molecules specifically compartmentalized inside the cells, in order to “feel” from the inside the cellular structure constraining the diffusion of these molecules. Diffusion data are then modeled, by numerical simulations in “virtual cells” massively performed on graphics processing unit, to extract the complexity and size of the cells in the investigated brain volume. The proposed strategy, although still demanding further validation, let us foresee the possibility to non-invasively measure brain cells morphology.