A spatial imaging-transcriptomics paradigm for deciphering the molecular basis of microscopic MRI in the normal brain and Alzheimer's disease.

Imaging genomics offers powerful links between genetics and neuroimaging phenotypes, but conventional methods often lack spatial correspondence. We introduce spatial imaging-transcriptomics, a paradigm that directly integrates magnetic resonance imaging (MRI) and spatially resolved transcriptomics (ST) with matched resolution from the same brain through spatially co-registered maps. We apply this framework to wild-type and Alzheimer's disease (AD) mouse models to uncover the molecular underpinnings of diffusion MRI. In the normal brain, fractional anisotropy (FA) correlates with myelination processes and oligodendrocytes, whereas diffusivity indices are associated with neurons. Furthermore, we identified the genetic basis for observed cortical gradients in diffusion tensor imaging (DTI). In the AD mouse model, reduced FA is linked to myelin and oligodendrocyte change and β-amyloid deposition. Critically, imaging-defined disease foci can guide differential gene expression analysis to uncover AD-specific genetic alterations. The spatial imaging-transcriptomics paradigm provides an unprecedented high-resolution bridge between neuroimaging and transcriptomics, benefitting the decipherment of complex molecular events in brain disorders.