Diffusion MRI has the ability to reveal the complex connectivity of the human brain. However, the link between the diffusion signal and the underlying tissue microstructure remains elusive. To drive diffusion MRI validation, we present BigMac: a unique dataset which combines ultra-high angular resolution diffusion MRI with microscopy throughout an adult macaque brain.
With this dataset we ask how under-sampling q-space biases our reconstruction of the ‘true’ diffusion profile. Our results indicate that the error associated with interpolating under-sampled data decays exponentially with the angular resolution of sampling, and that high angular resolution is necessary to characterise acutely crossing fibres.