Topological principles and developmental algorithms might refine diffusion tractography

Abstract

The identification and reconstruction of axonal pathways in the living brain or "ex-vivo" is promising a revolution in connectivity studies bridging the gap from animal to human neuroanatomy with extensions to brain structural-functional correlates. Unfortunately, the methods suffer from juvenile drawbacks. In this perspective paper we mention several computational and developmental principles, which might stimulate a new generation of algorithms and a discussion bridging the neuroimaging and neuroanatomy communities.

Keywords: Axons; Brain development; Brain pathways; Diffusion MRI; Tractography.

Fig. 1

The topological organization of the entire macaque cortical processing system as then known. A total of 758 connections between the 73 areas are represented, of which 136 (18%) are one-way. This connectivity represents 15% of the possible connections between these areas. This non-arbitrary structure represents in a spatial framework the organizational structure of the network of cortico-cortical connections of this animal. For the explanation of symbols see Young (1993). An upgraded version of the same figure exists in Young et al. (1995) (from Young , modified)

Fig. 2

Axons are organized in tight fascicles in their initial trajectory. a Shows BDA labeled axons originating from an injection site near the areas 9/46 border in a macaque. b, c show enlarged views of the axonal fascicle. In c some axonal segments are down for clarity. The axons defasciculate further down along their course (see Fig. 1 in Caminiti et al. 2009)

Fig. 3

Schematic representation of radial glia at the time of neuronal migration and of axonal ingrowth. The left part of the figure is borrowed from Rakic (1972). The right part of the figure is reproduced from Norris and Kalil (1991)

Fig. 4

The cortical flow algorithm reproduces to some extent the growth of axons into the gray matter in early development. a Shows the initial gyration, b the gyration after applying the cortical flow algorithm, c is the initial stage of streamline ingrowth, d the final stage of streamline ingrowth. See also Online animation: cortical flow.gif