An image processing method for characterizing diffusivity in brain’s parenchyma: a case study of significantly non-uniform structures

Abstract

We present a simple method, based on the principle of diffusion imaged processing, for revealing the type and quantitative characteristic of transport processes in brain tissue from fluorescent microscopy data. We consider a general form of the diffusion kernel, which covers both normal and hindered anomalous diffusion in media with complex microstructure as well as its practical correlation-based implementation. Our method overcomes restrictions of the conventional methods, which require extracting small areas with a high degree of symmetry and allows to process the large scale images, when solute leakage sources are distributed along the branched vascular structure.